=== 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: 6b3d50053968d9441899c157df378c42aede3ea5 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/01 job id: 44643929 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/02 job id: 44643930 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/03 job id: 44643931 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/04 job id: 44643932 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/05 job id: 44643936 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/06 job id: 44643939 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/07 job id: 44643942 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/08 job id: 44643945 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/09 job id: 44643948 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/10 job id: 44643952 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/11 job id: 44643957 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/12 job id: 44643959 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/13 job id: 44643961 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/14 job id: 44643963 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/15 job id: 44643964 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/16 job id: 44643966 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/17 job id: 44643967 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/18 job id: 44643968 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/19 job id: 44643969 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/20 job id: 44643970 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/21 job id: 44643971 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/22 job id: 44643972 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/23 job id: 44643976 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/24 job id: 44643979 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/25 job id: 44643983 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/26 job id: 44643986 --- 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/6b3d50053968d9441899c157df378c42aede3ea5_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 0.021 0.033 134.345 134.346 farming_run 1 2.0 133.865 133.866 134.318 134.321 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459372E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.092 115.541 115.542 qs_energies 1 2.0 0.000 0.000 115.285 115.287 mp2_main 1 3.0 0.000 0.000 113.402 113.404 mp2_gpw_main 1 4.0 0.020 0.026 112.583 112.585 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.990 94.068 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.677 55.753 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 41.709 46.663 get_2c_integrals 1 6.0 0.000 0.000 37.587 38.139 integrate_v_rspace 273 8.0 0.439 0.451 25.100 29.739 pw_transfer 6555 10.6 0.370 0.386 27.394 27.818 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.031 26.230 grid_integrate_task_list 273 9.0 20.920 25.997 20.920 25.997 fft_wrap_pw1pw2_100 2178 12.4 1.222 1.433 23.550 23.774 compute_2c_integrals 1 7.0 0.003 0.003 19.801 19.802 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.929 19.392 mp2_eri_2c_integrate_gpw 1 9.0 2.361 2.430 18.925 19.390 rpa_ri_compute_en 1 5.0 0.007 0.009 18.488 18.544 cp_fm_cholesky_decompose 12 8.2 17.736 18.267 17.736 18.267 cholesky_decomp 1 7.0 0.000 0.000 16.638 17.176 fft3d_s 5443 13.4 16.097 16.278 16.119 16.302 ao_to_mo_and_store_B_mult_1 272 7.0 10.854 15.578 10.854 15.578 calculate_wavefunction 272 8.0 5.417 5.494 12.521 13.194 rpa_num_int 1 6.0 0.001 0.017 10.513 10.513 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.456 10.467 calc_mat_Q 8 8.0 0.000 0.000 9.311 9.412 contract_S_to_Q 8 9.0 0.000 0.000 8.735 8.832 calc_potential_gpw 544 9.5 0.005 0.006 8.304 8.795 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.262 8.566 parallel_gemm_fm 14 9.1 0.000 0.000 8.311 8.426 parallel_gemm_fm_cosma 14 10.1 8.311 8.426 8.311 8.426 potential_pw2rs 545 10.0 0.108 0.110 7.684 8.310 create_integ_mat 1 6.0 0.022 0.027 7.794 7.795 collocate_single_gaussian 272 10.0 0.039 0.042 7.481 7.730 array2fm 1 7.0 0.000 0.000 6.730 7.301 pw_scatter_s 2720 13.7 4.426 4.591 4.426 4.591 pw_gather_s 2722 13.2 3.874 4.307 3.874 4.307 array2fm_buffer_send 1 8.0 2.992 3.158 2.992 3.158 pw_poisson_solve 545 10.5 1.130 1.205 2.202 2.532 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.392 1.535 2.342 2.528 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.585234, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 0.027 0.039 397.628 397.630 farming_run 1 2.0 396.795 396.800 397.592 397.595 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.228042E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 25009875. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 210.138 210.138 qs_energies 1 2.0 0.000 0.000 209.635 209.853 scf_env_do_scf 1 3.0 0.000 0.000 106.502 106.502 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.622 105.631 rebuild_ks_matrix 4 6.0 0.000 0.000 105.621 105.630 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.621 105.629 hfx_ks_matrix 4 8.0 0.001 0.001 105.233 105.237 integrate_four_center 4 9.0 0.144 0.457 105.233 105.236 mp2_main 1 3.0 0.002 0.020 102.845 103.063 mp2_gpw_main 1 4.0 0.050 0.172 101.985 102.186 integrate_four_center_main 4 10.0 0.135 0.623 96.776 99.409 integrate_four_center_bin 265 11.0 96.641 98.972 96.641 98.972 init_scf_loop 1 4.0 0.000 0.000 92.360 92.361 mp2_ri_gpw_compute_in 1 5.0 0.068 0.081 74.874 75.925 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.418 55.478 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.156 42.090 47.126 integrate_v_rspace 95 8.0 0.398 0.571 28.473 33.303 pw_transfer 2240 10.6 0.143 0.160 29.921 30.376 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.940 29.401 grid_integrate_task_list 95 9.0 23.762 28.818 23.762 28.818 mp2_ri_gpw_compute_en 1 5.0 0.171 0.315 26.933 28.612 ao_to_mo_and_store_B_mult_1 91 7.0 10.639 27.112 10.639 27.112 fft_wrap_pw1pw2_100 730 12.4 1.260 1.441 26.636 27.057 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.919 25.030 25.040 get_2c_integrals 1 6.0 0.006 0.037 20.360 20.393 compute_2c_integrals 1 7.0 0.011 0.035 19.330 19.344 compute_2c_integrals_loop_lm 1 8.0 0.003 0.010 18.856 19.159 mp2_eri_2c_integrate_gpw 1 9.0 1.747 1.858 18.853 19.158 fft3d_s 1823 13.4 18.414 18.758 18.428 18.771 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.140 14.140 calculate_wavefunction 91 8.0 2.015 2.049 9.738 9.937 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.576 8.773 9.518 potential_pw2rs 186 10.0 0.033 0.035 8.636 9.212 local_gemm 172 8.0 8.218 8.943 8.218 8.943 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.009 8.235 8.681 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.529 8.018 8.397 collocate_single_gaussian 91 10.0 0.017 0.026 7.872 8.187 calc_potential_gpw 182 9.5 0.002 0.002 7.905 8.173 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.530 6.345 6.530 mp_sendrecv_dm3 2068 8.0 6.054 6.474 6.054 6.474 mp_sync 38 10.4 3.139 6.290 3.139 6.290 pw_gather_s 912 13.2 4.934 5.430 4.934 5.430 pw_scatter_s 910 13.7 3.934 4.231 3.934 4.231 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.081505, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 451.457024E+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 589622. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 54.749 54.751 qs_mol_dyn_low 1 2.0 0.003 0.003 54.030 54.039 qs_forces 11 3.9 0.002 0.002 53.968 53.968 qs_energies 11 4.9 0.001 0.002 52.481 52.499 scf_env_do_scf 11 5.9 0.000 0.001 46.081 46.081 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.892 43.893 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.015 35.506 qs_scf_new_mos 108 7.5 0.000 0.001 33.515 33.796 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.514 33.796 ot_scf_mini 108 9.5 0.002 0.002 31.782 31.949 multiply_cannon 2286 13.5 0.185 0.192 27.079 28.846 multiply_cannon_loop 2286 14.5 1.487 1.568 26.361 28.102 velocity_verlet 10 3.0 0.002 0.002 26.056 26.056 ot_mini 108 10.5 0.001 0.001 20.305 20.546 qs_ot_get_derivative 108 11.5 0.001 0.001 17.329 17.538 mp_waitall_1 245248 16.5 9.217 16.023 9.217 16.023 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.523 14.554 multiply_cannon_multrec 54864 15.5 4.204 6.486 8.044 11.843 rebuild_ks_matrix 119 8.3 0.000 0.000 8.092 8.244 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.030 8.092 8.244 multiply_cannon_sync_h2d 54864 15.5 5.842 7.853 5.842 7.853 mp_sum_l 7207 12.9 5.953 7.772 5.953 7.772 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.150 7.284 qs_ot_get_p 119 10.4 0.001 0.001 6.711 7.000 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.098 6.532 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.879 6.006 dbcsr_mm_accdrv_process 76910 16.1 1.162 1.840 3.762 5.465 init_scf_run 11 5.9 0.000 0.001 5.185 5.185 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.185 5.185 sum_up_and_integrate 119 10.3 0.012 0.014 4.588 4.594 integrate_v_rspace 119 11.3 0.002 0.002 4.575 4.583 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.302 4.438 calculate_rho_elec 119 8.7 0.011 0.017 4.302 4.438 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.832 3.913 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.981 3.704 jit_kernel_multiply 13 15.8 2.537 3.586 2.537 3.586 calculate_dm_sparse 119 9.5 0.000 0.001 3.085 3.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.888 3.210 apply_single 119 13.6 0.000 0.000 2.888 3.209 calculate_first_density_matrix 1 7.0 0.000 0.000 2.993 2.997 rs_pw_transfer 974 11.9 0.011 0.013 2.830 2.908 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.896 2.896 ot_diis_step 108 11.5 0.006 0.006 2.694 2.695 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.670 2.671 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.589 2.659 cp_fm_redistribute_end 50 14.0 2.430 2.648 2.435 2.649 cp_fm_diag_elpa_base 50 14.0 0.212 2.563 0.213 2.572 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.442 2.444 density_rs2pw 119 9.7 0.004 0.005 2.317 2.431 acc_transpose_blocks 54864 15.5 0.234 0.255 1.776 2.220 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.154 2.220 init_scf_loop 11 6.9 0.000 0.000 2.172 2.172 grid_integrate_task_list 119 12.3 2.026 2.133 2.026 2.133 wfi_extrapolate 11 7.9 0.001 0.001 2.127 2.127 mp_sum_d 4127 12.0 1.353 1.967 1.353 1.967 potential_pw2rs 119 12.3 0.004 0.004 1.936 1.948 pw_transfer 1439 11.6 0.053 0.059 1.869 1.943 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.792 1.870 make_m2s 4572 13.5 0.054 0.057 1.673 1.728 make_images 4572 14.5 0.132 0.137 1.590 1.646 fft3d_ps 1201 14.6 0.368 0.474 1.563 1.636 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.520 1.541 mp_alltoall_d11v 2130 13.8 1.295 1.482 1.295 1.482 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.384 1.466 mp_waitany 12084 13.8 1.248 1.446 1.248 1.446 grid_collocate_task_list 119 9.7 1.290 1.373 1.290 1.373 parallel_gemm_fm 81 9.0 0.000 0.000 1.149 1.153 parallel_gemm_fm_cosma 81 10.0 1.149 1.153 1.149 1.153 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.738 1.121 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.751000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.382464E+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 651108. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.057 40.647 40.648 qs_mol_dyn_low 1 2.0 0.003 0.004 40.292 40.299 qs_forces 11 3.9 0.041 0.322 39.531 39.532 qs_energies 11 4.9 0.001 0.001 37.526 37.530 scf_env_do_scf 11 5.9 0.000 0.001 31.787 31.787 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.307 29.308 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.019 22.391 qs_scf_new_mos 108 7.5 0.001 0.001 19.995 20.234 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.994 20.234 ot_scf_mini 108 9.5 0.002 0.003 19.100 19.275 velocity_verlet 10 3.0 0.001 0.002 19.002 19.018 multiply_cannon 2286 13.5 0.207 0.215 16.929 18.519 multiply_cannon_loop 2286 14.5 0.900 0.968 15.842 17.328 ot_mini 108 10.5 0.001 0.001 11.891 12.131 mp_waitall_1 200699 16.5 5.728 10.960 5.728 10.960 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.220 9.758 qs_ot_get_derivative 108 11.5 0.009 0.197 9.426 9.607 multiply_cannon_multrec 27432 15.5 1.957 4.309 5.941 8.766 rebuild_ks_matrix 119 8.3 0.000 0.000 7.269 7.404 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.269 7.404 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.432 6.554 dbcsr_mm_accdrv_process 47894 16.0 3.141 5.439 3.915 5.943 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.607 4.632 qs_ot_get_p 119 10.4 0.001 0.001 4.331 4.572 init_scf_run 11 5.9 0.000 0.001 4.495 4.496 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.495 4.496 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.215 4.328 apply_single 119 13.6 0.000 0.000 3.215 4.328 sum_up_and_integrate 119 10.3 0.024 0.026 4.240 4.244 integrate_v_rspace 119 11.3 0.002 0.002 4.216 4.224 mp_sum_l 7207 12.9 2.195 4.189 2.195 4.189 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.867 3.900 calculate_rho_elec 119 8.7 0.021 0.024 3.867 3.899 rs_pw_transfer 974 11.9 0.010 0.011 2.668 3.109 calculate_first_density_matrix 1 7.0 0.000 0.000 2.977 2.980 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.900 2.919 multiply_cannon_sync_h2d 27432 15.5 2.195 2.891 2.195 2.891 make_m2s 4572 13.5 0.052 0.053 2.454 2.678 density_rs2pw 119 9.7 0.004 0.004 2.194 2.648 calculate_dm_sparse 119 9.5 0.000 0.001 2.553 2.627 make_images 4572 14.5 0.199 0.234 2.365 2.588 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.464 2.465 init_scf_loop 11 6.9 0.000 0.000 2.460 2.460 ot_diis_step 108 11.5 0.011 0.014 2.415 2.416 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.105 2.196 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.130 2.131 cp_fm_redistribute_end 50 14.0 1.763 2.106 1.767 2.107 pw_transfer 1439 11.6 0.065 0.072 2.016 2.059 cp_fm_diag_elpa_base 50 14.0 0.326 1.997 0.339 2.058 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.034 2.036 acc_transpose_blocks 27432 15.5 0.112 0.116 1.710 2.010 potential_pw2rs 119 12.3 0.006 0.006 1.967 1.978 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.924 1.971 grid_integrate_task_list 119 12.3 1.827 1.944 1.827 1.944 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.820 1.858 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.768 1.777 jit_kernel_multiply 8 16.1 0.722 1.719 0.722 1.719 fft3d_ps 1201 14.6 0.511 0.562 1.626 1.666 prepare_preconditioner 11 7.9 0.000 0.000 1.540 1.567 make_preconditioner 11 8.9 0.000 0.000 1.540 1.567 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.510 1.556 make_images_data 4572 15.5 0.044 0.051 1.124 1.509 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.444 1.501 wfi_extrapolate 11 7.9 0.001 0.001 1.465 1.465 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.986 1.414 acc_transpose_blocks_kernels 27432 16.5 0.183 0.273 1.172 1.387 grid_collocate_task_list 119 9.7 1.230 1.359 1.230 1.359 mp_sum_d 4127 12.0 0.863 1.357 0.863 1.357 mp_alltoall_d11v 2130 13.8 1.199 1.348 1.199 1.348 mp_allgather_i34 2286 14.5 0.531 1.288 0.531 1.288 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.159 1.204 jit_kernel_transpose 5 15.5 0.989 1.122 0.989 1.122 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.145 0.554 0.999 mp_waitany 5720 13.7 0.528 0.994 0.528 0.994 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.981 0.983 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.895 0.908 mp_alltoall_z22v 1201 16.6 0.720 0.821 0.720 0.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.648000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 520.949760E+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 590179. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 33.941 33.942 qs_mol_dyn_low 1 2.0 0.003 0.004 33.541 33.549 qs_forces 11 3.9 0.002 0.005 33.481 33.482 qs_energies 11 4.9 0.003 0.013 31.860 31.862 scf_env_do_scf 11 5.9 0.001 0.003 26.463 26.463 scf_env_do_scf_inner_loop 108 6.5 0.011 0.039 23.672 23.674 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.919 17.053 velocity_verlet 10 3.0 0.001 0.002 15.681 15.683 qs_scf_new_mos 108 7.5 0.001 0.001 15.214 15.231 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.213 15.231 ot_scf_mini 108 9.5 0.002 0.003 14.489 14.508 multiply_cannon 2286 13.5 0.195 0.203 13.400 14.121 multiply_cannon_loop 2286 14.5 0.637 0.665 12.585 13.336 ot_mini 108 10.5 0.001 0.001 8.968 8.984 multiply_cannon_multrec 18288 15.5 1.941 2.892 7.069 7.479 qs_ot_get_derivative 108 11.5 0.001 0.001 7.433 7.451 rebuild_ks_matrix 119 8.3 0.000 0.000 6.686 6.706 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.041 6.685 6.706 dbcsr_mm_accdrv_process 38222 16.0 4.309 5.711 5.045 6.005 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.909 5.927 sum_up_and_integrate 119 10.3 0.031 0.032 4.108 4.115 integrate_v_rspace 119 11.3 0.002 0.004 4.077 4.088 mp_waitall_1 158411 16.6 2.784 4.007 2.784 4.007 init_scf_run 11 5.9 0.000 0.001 3.954 3.954 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.953 3.954 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.568 3.576 calculate_rho_elec 119 8.7 0.030 0.031 3.568 3.575 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.763 3.474 qs_ot_get_p 119 10.4 0.001 0.002 3.363 3.384 rs_pw_transfer 974 11.9 0.009 0.010 2.574 2.852 init_scf_loop 11 6.9 0.002 0.008 2.772 2.776 calculate_first_density_matrix 1 7.0 0.001 0.008 2.667 2.667 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.460 2.525 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.149 2.507 apply_single 119 13.6 0.000 0.000 2.149 2.507 density_rs2pw 119 9.7 0.004 0.004 2.091 2.345 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.218 2.224 jit_kernel_multiply 10 16.2 0.684 2.202 0.684 2.202 calculate_dm_sparse 119 9.5 0.000 0.000 2.166 2.183 make_m2s 4572 13.5 0.044 0.045 1.864 2.023 pw_transfer 1439 11.6 0.066 0.071 1.975 1.987 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.946 1.946 make_images 4572 14.5 0.188 0.201 1.780 1.937 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.882 1.897 grid_integrate_task_list 119 12.3 1.798 1.883 1.798 1.883 potential_pw2rs 119 12.3 0.007 0.008 1.875 1.883 prepare_preconditioner 11 7.9 0.000 0.002 1.801 1.805 make_preconditioner 11 8.9 0.001 0.004 1.801 1.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.654 1.745 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.706 1.708 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.694 1.695 cp_fm_redistribute_end 50 14.0 1.262 1.672 1.263 1.672 cp_fm_diag_elpa_base 50 14.0 0.391 1.588 0.406 1.632 mp_sum_l 7207 12.9 1.278 1.628 1.278 1.628 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.612 1.618 fft3d_ps 1201 14.6 0.522 0.546 1.562 1.578 multiply_cannon_sync_h2d 18288 15.5 1.366 1.561 1.366 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.536 1.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.508 1.520 ot_diis_step 108 11.5 0.011 0.011 1.510 1.510 fft_wrap_pw1pw2_140 487 13.2 0.089 0.093 1.472 1.488 grid_collocate_task_list 119 9.7 1.215 1.373 1.215 1.373 acc_transpose_blocks 18288 15.5 0.077 0.079 1.256 1.363 wfi_extrapolate 11 7.9 0.001 0.001 1.235 1.235 qs_energies_init_hamiltonians 11 5.9 0.015 0.020 1.172 1.173 make_images_data 4572 15.5 0.044 0.048 0.841 1.007 mp_waitany 9880 13.7 0.643 0.951 0.643 0.951 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.727 0.938 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.893 0.916 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.814 0.916 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.616 0.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.852 mp_alltoall_d11v 2130 13.8 0.685 0.840 0.685 0.840 cp_fm_cholesky_invert 11 10.9 0.768 0.772 0.768 0.772 qs_env_update_s_mstruct 11 6.9 0.002 0.011 0.698 0.770 mp_alltoall_z22v 1201 16.6 0.637 0.755 0.637 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.664 0.744 jit_kernel_transpose 5 15.6 0.603 0.699 0.603 0.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.942000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 550.236160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.062 36.249 36.251 qs_mol_dyn_low 1 2.0 0.004 0.007 35.893 35.901 qs_forces 11 3.9 0.002 0.004 35.833 35.834 qs_energies 11 4.9 0.031 0.070 34.093 34.100 scf_env_do_scf 11 5.9 0.001 0.002 28.441 28.442 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.808 24.808 dbcsr_multiply_generic 2286 12.5 0.102 0.104 18.329 18.427 velocity_verlet 10 3.0 0.001 0.002 18.013 18.015 qs_scf_new_mos 108 7.5 0.001 0.001 16.214 16.262 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.213 16.262 ot_scf_mini 108 9.5 0.003 0.003 15.281 15.335 multiply_cannon 2286 13.5 0.225 0.253 14.728 14.984 multiply_cannon_loop 2286 14.5 0.939 0.965 13.786 14.141 multiply_cannon_multrec 27432 15.5 2.330 3.005 8.939 9.311 ot_mini 108 10.5 0.001 0.001 9.206 9.276 dbcsr_mm_accdrv_process 47916 15.9 5.373 7.304 6.514 7.704 qs_ot_get_derivative 108 11.5 0.001 0.001 7.393 7.449 rebuild_ks_matrix 119 8.3 0.000 0.000 6.758 6.806 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.030 6.758 6.805 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.010 6.051 init_scf_run 11 5.9 0.000 0.001 4.114 4.117 scf_env_initial_rho_setup 11 6.9 0.005 0.015 4.114 4.117 sum_up_and_integrate 119 10.3 0.036 0.040 3.858 3.867 integrate_v_rspace 119 11.3 0.002 0.004 3.823 3.832 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.652 3.687 calculate_rho_elec 119 8.7 0.040 0.046 3.651 3.687 init_scf_loop 11 6.9 0.005 0.038 3.613 3.613 qs_ot_get_p 119 10.4 0.001 0.001 3.467 3.541 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.769 3.166 calculate_first_density_matrix 1 7.0 0.001 0.006 2.721 2.726 prepare_preconditioner 11 7.9 0.000 0.002 2.634 2.648 make_preconditioner 11 8.9 0.005 0.037 2.634 2.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.232 2.561 calculate_dm_sparse 119 9.5 0.000 0.000 2.395 2.447 mp_waitall_1 137007 16.6 1.789 2.376 1.789 2.376 rs_pw_transfer 974 11.9 0.009 0.010 2.127 2.333 make_m2s 4572 13.5 0.054 0.056 2.205 2.307 density_rs2pw 119 9.7 0.004 0.004 2.085 2.268 make_images 4572 14.5 0.269 0.326 2.098 2.198 pw_transfer 1439 11.6 0.066 0.071 2.155 2.194 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.047 2.170 apply_single 119 13.6 0.000 0.000 2.047 2.170 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.123 2.132 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.062 2.106 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.928 1.958 grid_integrate_task_list 119 12.3 1.823 1.908 1.823 1.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.887 1.889 jit_kernel_multiply 8 15.8 1.081 1.852 1.081 1.852 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.806 1.806 fft3d_ps 1201 14.6 0.558 0.613 1.736 1.773 ot_diis_step 108 11.5 0.012 0.012 1.772 1.772 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.668 1.718 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.630 potential_pw2rs 119 12.3 0.008 0.010 1.618 1.622 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.580 1.590 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.546 1.547 cp_fm_redistribute_end 50 14.0 1.024 1.521 1.025 1.521 acc_transpose_blocks 27432 15.5 0.115 0.117 1.463 1.500 cp_fm_diag_elpa_base 50 14.0 0.473 1.452 0.493 1.490 multiply_cannon_metrocomm3 27432 15.5 0.038 0.038 0.738 1.372 wfi_extrapolate 11 7.9 0.001 0.001 1.339 1.339 grid_collocate_task_list 119 9.7 1.222 1.328 1.222 1.328 qs_energies_init_hamiltonians 11 5.9 0.010 0.071 1.194 1.253 mp_sum_l 7207 12.9 0.974 1.252 0.974 1.252 dbcsr_complete_redistribute 329 12.2 0.148 0.201 0.895 1.173 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.119 cp_fm_upper_to_full 72 13.5 0.801 1.112 0.801 1.112 multiply_cannon_sync_h2d 27432 15.5 0.994 1.070 0.994 1.070 make_images_data 4572 15.5 0.045 0.049 0.855 0.973 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.667 0.939 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.737 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.874 mp_alltoall_d11v 2130 13.8 0.724 0.865 0.724 0.865 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.833 0.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.844 mp_alltoall_z22v 1201 16.6 0.763 0.798 0.763 0.798 cp_fm_cholesky_invert 11 10.9 0.780 0.783 0.780 0.783 mp_alltoall_i22 627 13.8 0.438 0.735 0.438 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.251000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.545455, yerr=2.387986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 600.342528E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 28.426 28.427 qs_mol_dyn_low 1 2.0 0.003 0.003 28.262 28.269 qs_forces 11 3.9 0.002 0.002 28.190 28.191 qs_energies 11 4.9 0.001 0.001 26.467 26.469 scf_env_do_scf 11 5.9 0.000 0.001 21.750 21.750 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.191 19.192 velocity_verlet 10 3.0 0.002 0.002 14.602 14.606 dbcsr_multiply_generic 2286 12.5 0.119 0.123 12.163 12.234 qs_scf_new_mos 108 7.5 0.001 0.001 11.042 11.063 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.041 11.062 ot_scf_mini 108 9.5 0.002 0.002 10.367 10.391 multiply_cannon 2286 13.5 0.231 0.241 9.576 10.040 multiply_cannon_loop 2286 14.5 0.330 0.339 8.664 8.838 rebuild_ks_matrix 119 8.3 0.000 0.000 6.069 6.085 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.068 6.085 multiply_cannon_multrec 9144 15.5 1.646 1.891 5.792 6.008 ot_mini 108 10.5 0.001 0.001 5.754 5.782 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.411 5.426 qs_ot_get_derivative 108 11.5 0.001 0.001 4.431 4.457 dbcsr_mm_accdrv_process 12550 15.8 3.162 4.061 4.045 4.120 sum_up_and_integrate 119 10.3 0.038 0.042 3.712 3.717 integrate_v_rspace 119 11.3 0.003 0.003 3.674 3.679 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.620 3.628 calculate_rho_elec 119 8.7 0.059 0.061 3.619 3.627 init_scf_run 11 5.9 0.000 0.001 3.270 3.270 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.270 3.270 qs_ot_get_p 119 10.4 0.001 0.001 2.831 2.867 init_scf_loop 11 6.9 0.000 0.000 2.538 2.539 pw_transfer 1439 11.6 0.066 0.069 2.199 2.210 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.105 2.119 calculate_first_density_matrix 1 7.0 0.000 0.000 2.113 2.114 density_rs2pw 119 9.7 0.004 0.004 1.934 2.077 mp_waitall_1 115863 16.7 1.462 1.994 1.462 1.994 grid_integrate_task_list 119 12.3 1.853 1.923 1.853 1.923 make_m2s 4572 13.5 0.034 0.035 1.750 1.906 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.896 1.898 rs_pw_transfer 974 11.9 0.008 0.008 1.696 1.839 make_images 4572 14.5 0.267 0.300 1.661 1.815 prepare_preconditioner 11 7.9 0.000 0.000 1.788 1.792 make_preconditioner 11 8.9 0.000 0.000 1.788 1.792 fft3d_ps 1201 14.6 0.562 0.572 1.764 1.775 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.731 1.747 jit_kernel_multiply 10 15.7 0.845 1.739 0.845 1.739 calculate_dm_sparse 119 9.5 0.000 0.000 1.699 1.718 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.699 1.699 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.672 1.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.587 1.588 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.498 1.516 potential_pw2rs 119 12.3 0.010 0.010 1.437 1.442 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.409 1.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.406 1.406 grid_collocate_task_list 119 9.7 1.275 1.393 1.275 1.393 cp_fm_redistribute_end 50 14.0 0.701 1.382 0.702 1.383 cp_fm_diag_elpa_base 50 14.0 0.636 1.313 0.680 1.368 ot_diis_step 108 11.5 0.012 0.013 1.299 1.299 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.237 1.238 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.176 1.194 apply_single 119 13.6 0.000 0.000 1.176 1.193 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.146 1.158 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.119 1.123 wfi_extrapolate 11 7.9 0.001 0.001 1.107 1.107 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.792 1.024 make_images_data 4572 15.5 0.039 0.042 0.813 0.988 acc_transpose_blocks 9144 15.5 0.039 0.039 0.907 0.924 cp_fm_cholesky_invert 11 10.9 0.919 0.922 0.919 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.920 mp_alltoall_d11v 2130 13.8 0.799 0.894 0.799 0.894 multiply_cannon_sync_h2d 9144 15.5 0.715 0.785 0.715 0.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.783 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.692 0.740 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.721 0.731 mp_alltoall_z22v 1201 16.6 0.650 0.701 0.650 0.701 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.376 0.685 acc_transpose_blocks_kernels 9144 16.5 0.118 0.122 0.665 0.679 mp_allgather_i34 2286 14.5 0.245 0.674 0.245 0.674 yz_to_x 606 15.1 0.265 0.275 0.603 0.619 mp_sum_l 7207 12.9 0.380 0.597 0.380 0.597 x_to_yz 595 16.2 0.277 0.289 0.588 0.595 mp_waitany 5200 13.7 0.441 0.579 0.441 0.579 qs_create_task_list 11 7.9 0.011 0.012 0.554 0.577 dbcsr_complete_redistribute 329 12.2 0.169 0.183 0.548 0.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.427000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.181818, yerr=5.874725 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 739.188736E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 42.681 42.682 qs_mol_dyn_low 1 2.0 0.003 0.003 42.437 42.444 qs_forces 11 3.9 0.002 0.003 42.098 42.099 qs_energies 11 4.9 0.001 0.001 40.098 40.102 scf_env_do_scf 11 5.9 0.001 0.001 33.995 33.995 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.993 25.994 velocity_verlet 10 3.0 0.259 0.265 23.604 23.613 dbcsr_multiply_generic 2286 12.5 0.125 0.127 18.157 18.323 qs_scf_new_mos 108 7.5 0.001 0.001 16.148 16.243 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.147 16.242 multiply_cannon 2286 13.5 0.302 0.311 14.292 15.183 ot_scf_mini 108 9.5 0.002 0.002 15.076 15.174 multiply_cannon_loop 2286 14.5 0.344 0.350 13.028 13.947 ot_mini 108 10.5 0.001 0.001 9.204 9.318 multiply_cannon_multrec 9144 15.5 3.395 4.840 8.727 8.883 init_scf_loop 11 6.9 0.000 0.000 7.975 7.977 rebuild_ks_matrix 119 8.3 0.000 0.000 7.516 7.658 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.515 7.657 qs_ot_get_derivative 108 11.5 0.001 0.001 7.178 7.279 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.808 6.938 prepare_preconditioner 11 7.9 0.000 0.000 6.848 6.861 make_preconditioner 11 8.9 0.000 0.000 6.848 6.861 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.385 6.742 dbcsr_mm_accdrv_process 12550 15.8 4.246 5.887 5.207 6.644 cp_fm_upper_to_full 72 14.2 3.206 4.611 3.206 4.611 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.265 4.271 calculate_rho_elec 119 8.7 0.118 0.121 4.265 4.270 init_scf_run 11 5.9 0.000 0.001 4.061 4.061 scf_env_initial_rho_setup 11 6.9 0.003 0.004 4.061 4.061 sum_up_and_integrate 119 10.3 0.066 0.067 4.040 4.046 integrate_v_rspace 119 11.3 0.003 0.004 3.974 3.981 mp_waitall_1 94719 16.7 2.622 3.700 2.622 3.700 qs_ot_get_p 119 10.4 0.001 0.001 3.150 3.284 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.827 3.281 dbcsr_complete_redistribute 329 12.2 0.294 0.299 1.989 2.838 pw_transfer 1439 11.6 0.068 0.069 2.815 2.820 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.716 2.721 calculate_first_density_matrix 1 7.0 0.000 0.000 2.612 2.614 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.551 2.607 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.222 2.593 apply_single 119 13.6 0.000 0.000 2.222 2.593 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.691 2.524 calculate_dm_sparse 119 9.5 0.000 0.000 2.464 2.482 make_m2s 4572 13.5 0.038 0.038 2.218 2.397 mp_alltoall_i22 627 13.8 1.496 2.360 1.496 2.360 fft3d_ps 1201 14.6 0.593 0.603 2.340 2.344 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.297 2.305 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.458 2.285 make_images 4572 14.5 0.352 0.384 2.096 2.275 density_rs2pw 119 9.7 0.004 0.004 2.196 2.214 grid_integrate_task_list 119 12.3 2.057 2.069 2.057 2.069 ot_diis_step 108 11.5 0.014 0.014 2.001 2.002 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.987 1.988 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.914 1.914 mp_sum_l 7207 12.9 1.082 1.857 1.082 1.857 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 1.780 1.781 jit_kernel_multiply 8 15.5 0.935 1.738 0.935 1.738 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.674 1.725 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.669 1.698 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.644 1.645 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.538 1.546 grid_collocate_task_list 119 9.7 1.487 1.507 1.487 1.507 rs_pw_transfer 974 11.9 0.009 0.009 1.430 1.451 potential_pw2rs 119 12.3 0.014 0.015 1.438 1.441 cp_fm_cholesky_invert 11 10.9 1.431 1.434 1.431 1.434 wfi_extrapolate 11 7.9 0.001 0.001 1.378 1.378 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.355 cp_fm_diag_elpa_base 50 14.0 1.211 1.264 1.353 1.353 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.089 1.341 make_images_data 4572 15.5 0.043 0.046 1.035 1.256 acc_transpose_blocks 9144 15.5 0.039 0.040 1.143 1.255 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.122 1.141 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.110 1.131 mp_alltoall_d11v 2130 13.8 1.060 1.087 1.060 1.087 multiply_cannon_sync_h2d 9144 15.5 1.040 1.044 1.040 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.987 1.038 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 0.892 1.004 qs_create_task_list 11 7.9 0.000 0.001 0.939 0.951 generate_qs_task_list 11 8.9 0.369 0.388 0.938 0.951 yz_to_x 606 15.1 0.460 0.475 0.937 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.932 jit_kernel_transpose 5 15.6 0.774 0.887 0.774 0.887 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.682000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=692.545455, yerr=13.124326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 500.191232E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.048 84.040 84.042 qs_mol_dyn_low 1 2.0 0.003 0.003 83.776 83.786 qs_forces 11 3.9 0.002 0.002 83.707 83.708 qs_energies 11 4.9 0.001 0.002 80.848 80.861 scf_env_do_scf 11 5.9 0.001 0.004 71.641 71.643 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.076 66.076 dbcsr_multiply_generic 2055 12.4 0.106 0.109 52.700 52.983 qs_scf_new_mos 99 7.5 0.000 0.001 48.459 48.581 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.458 48.581 ot_scf_mini 99 9.5 0.002 0.002 46.040 46.208 multiply_cannon 2055 13.4 0.179 0.186 43.086 44.067 multiply_cannon_loop 2055 14.4 1.549 1.584 42.108 43.056 velocity_verlet 10 3.0 0.002 0.002 42.228 42.229 ot_mini 99 10.5 0.001 0.001 27.830 27.989 qs_ot_get_derivative 99 11.5 0.001 0.001 20.987 21.151 multiply_cannon_multrec 49320 15.4 12.138 12.776 17.512 18.303 rebuild_ks_matrix 110 8.3 0.000 0.000 14.757 14.904 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.756 14.904 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.960 13.091 mp_waitall_1 220248 16.4 11.785 12.814 11.785 12.814 multiply_cannon_sync_h2d 49320 15.4 9.909 10.524 9.909 10.524 qs_ot_get_p 110 10.4 0.001 0.001 9.661 9.851 multiply_cannon_metrocomm3 49320 15.4 0.078 0.083 7.113 8.537 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.170 7.711 apply_single 110 13.6 0.000 0.001 7.170 7.710 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.132 7.672 sum_up_and_integrate 110 10.3 0.036 0.043 7.195 7.211 init_scf_run 11 5.9 0.000 0.001 7.185 7.185 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.185 7.185 integrate_v_rspace 110 11.3 0.003 0.003 7.158 7.183 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.856 6.945 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.755 6.914 calculate_rho_elec 110 8.6 0.021 0.025 6.754 6.913 ot_diis_step 99 11.5 0.005 0.006 6.554 6.554 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.428 6.474 mp_sum_l 6514 12.8 5.248 6.156 5.248 6.156 dbcsr_mm_accdrv_process 87628 16.1 2.073 2.179 5.251 5.574 init_scf_loop 11 6.9 0.000 0.001 5.535 5.536 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.422 5.422 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.936 4.937 cp_fm_redistribute_end 48 14.0 4.308 4.910 4.312 4.911 cp_fm_diag_elpa_base 48 14.0 0.593 4.759 0.597 4.791 rs_pw_transfer 902 11.9 0.012 0.013 3.734 4.320 calculate_dm_sparse 110 9.5 0.000 0.001 4.151 4.244 density_rs2pw 110 9.6 0.004 0.005 3.531 4.137 wfi_extrapolate 11 7.9 0.001 0.001 4.044 4.044 make_m2s 4110 13.4 0.061 0.066 3.922 4.032 make_images 4110 14.4 0.176 0.190 3.827 3.939 multiply_cannon_metrocomm1 49320 15.4 0.060 0.065 2.739 3.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.633 3.637 pw_transfer 1331 11.6 0.055 0.070 3.363 3.441 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.311 3.394 grid_integrate_task_list 110 12.3 3.247 3.393 3.247 3.393 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.274 3.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.307 3.348 prepare_preconditioner 11 7.9 0.000 0.000 3.297 3.316 make_preconditioner 11 8.9 0.000 0.000 3.297 3.316 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.072 3.119 calculate_first_density_matrix 1 7.0 0.000 0.000 3.045 3.056 jit_kernel_multiply 13 15.9 2.895 2.933 2.895 2.933 fft3d_ps 1111 14.6 0.775 0.881 2.803 2.872 fft_wrap_pw1pw2_140 451 13.1 0.171 0.191 2.768 2.855 potential_pw2rs 110 12.3 0.006 0.007 2.724 2.750 mp_waitany 14300 13.8 1.806 2.523 1.806 2.523 mp_alltoall_d11v 2046 13.8 2.104 2.444 2.104 2.444 acc_transpose_blocks 49320 15.4 0.227 0.238 2.213 2.300 grid_collocate_task_list 110 9.6 2.083 2.292 2.083 2.292 mp_sum_d 3881 11.9 1.483 2.057 1.483 2.057 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.017 2.046 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.958 make_images_data 4110 15.4 0.042 0.046 1.816 1.951 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.571 1.850 cp_fm_cholesky_invert 11 10.9 1.809 1.813 1.809 1.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.042000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.636364, yerr=2.422348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 591.273984E+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 1294336. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.044 70.067 70.069 qs_mol_dyn_low 1 2.0 0.025 0.178 69.551 69.560 qs_forces 11 3.9 0.002 0.002 69.256 69.257 qs_energies 11 4.9 0.001 0.002 65.899 65.903 scf_env_do_scf 11 5.9 0.000 0.001 56.545 56.549 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 48.986 48.986 dbcsr_multiply_generic 2055 12.4 0.113 0.117 38.374 38.591 velocity_verlet 10 3.0 0.002 0.004 36.310 36.312 multiply_cannon 2055 13.4 0.221 0.243 31.812 33.176 qs_scf_new_mos 99 7.5 0.001 0.001 32.495 32.642 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.495 32.641 multiply_cannon_loop 2055 14.4 0.928 0.958 30.516 31.530 ot_scf_mini 99 9.5 0.003 0.003 30.841 30.982 ot_mini 99 10.5 0.001 0.001 18.122 18.259 multiply_cannon_multrec 24660 15.4 7.635 9.352 14.052 15.854 rebuild_ks_matrix 110 8.3 0.000 0.001 13.930 14.087 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.929 14.086 qs_ot_get_derivative 99 11.5 0.001 0.001 12.327 12.465 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.268 12.414 mp_waitall_1 176588 16.5 7.722 10.515 7.722 10.515 multiply_cannon_sync_h2d 24660 15.4 7.075 8.228 7.075 8.228 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.231 7.730 init_scf_loop 11 6.9 0.000 0.001 7.524 7.525 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.492 7.249 apply_single 110 13.6 0.000 0.001 6.491 7.248 init_scf_run 11 5.9 0.000 0.001 6.994 6.995 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.994 6.995 dbcsr_mm_accdrv_process 52282 16.1 4.881 5.651 6.256 6.719 sum_up_and_integrate 110 10.3 0.052 0.060 6.687 6.700 integrate_v_rspace 110 11.3 0.002 0.002 6.635 6.650 qs_ot_get_p 110 10.4 0.001 0.001 6.094 6.272 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.226 6.238 calculate_rho_elec 110 8.6 0.039 0.047 6.226 6.237 ot_diis_step 99 11.5 0.010 0.010 5.750 5.750 prepare_preconditioner 11 7.9 0.000 0.000 5.481 5.508 make_preconditioner 11 8.9 0.000 0.000 5.481 5.508 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.660 5.400 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.068 5.225 make_m2s 4110 13.4 0.057 0.060 4.142 4.639 make_images 4110 14.4 0.396 0.440 4.033 4.525 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.216 4.235 pw_transfer 1331 11.6 0.066 0.075 3.791 3.937 density_rs2pw 110 9.6 0.004 0.004 3.345 3.867 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.684 3.833 calculate_dm_sparse 110 9.5 0.001 0.001 3.778 3.810 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.791 3.792 wfi_extrapolate 11 7.9 0.001 0.001 3.539 3.539 rs_pw_transfer 902 11.9 0.012 0.013 2.967 3.507 calculate_first_density_matrix 1 7.0 0.000 0.001 3.365 3.368 fft_wrap_pw1pw2_140 451 13.1 0.203 0.221 3.144 3.295 grid_integrate_task_list 110 12.3 3.146 3.289 3.146 3.289 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.263 3.265 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.257 3.258 cp_fm_redistribute_end 48 14.0 2.441 3.239 2.443 3.239 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.152 3.234 cp_fm_diag_elpa_base 48 14.0 0.762 3.098 0.793 3.183 fft3d_ps 1111 14.6 1.101 1.317 3.034 3.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.894 2.962 make_images_data 4110 15.4 0.046 0.050 2.327 2.827 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.052 2.753 jit_kernel_multiply 12 16.3 1.027 2.670 1.027 2.670 cp_fm_cholesky_invert 11 10.9 2.534 2.541 2.534 2.541 potential_pw2rs 110 12.3 0.008 0.008 2.471 2.481 grid_collocate_task_list 110 9.6 2.053 2.478 2.053 2.478 mp_sum_l 6514 12.8 1.817 2.411 1.817 2.411 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.355 2.365 mp_alltoall_d11v 2046 13.8 1.765 2.039 1.765 2.039 acc_transpose_blocks 24660 15.4 0.113 0.117 1.968 1.997 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.934 1.953 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.859 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.825 1.840 mp_waitany 10164 13.8 1.221 1.734 1.221 1.734 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.766 1.720 mp_allgather_i34 2055 14.4 0.625 1.648 0.625 1.648 mp_irecv_dv 57340 16.2 0.642 1.600 0.642 1.600 rs_pw_transfer_RS2PW_140 121 11.5 0.204 0.214 0.956 1.488 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.373 1.487 dbcsr_complete_redistribute 325 12.2 0.244 0.308 1.135 1.405 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.069000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.909091, yerr=8.317064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 660.111360E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.084 61.611 61.614 qs_mol_dyn_low 1 2.0 0.003 0.004 60.560 60.570 qs_forces 11 3.9 0.015 0.027 60.494 60.495 qs_energies 11 4.9 0.001 0.005 57.231 57.247 scf_env_do_scf 11 5.9 0.001 0.002 48.887 48.888 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.222 40.224 velocity_verlet 10 3.0 0.001 0.002 32.647 32.648 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.663 28.921 qs_scf_new_mos 99 7.5 0.001 0.001 25.062 25.138 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.061 25.138 ot_scf_mini 99 9.5 0.003 0.004 23.851 23.948 multiply_cannon 2055 13.4 0.212 0.230 22.373 23.565 multiply_cannon_loop 2055 14.4 0.616 0.630 21.169 22.326 ot_mini 99 10.5 0.001 0.001 13.746 13.845 rebuild_ks_matrix 110 8.3 0.000 0.000 12.522 12.643 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.025 12.521 12.643 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.034 11.144 mp_waitall_1 139946 16.5 7.263 10.683 7.263 10.683 multiply_cannon_multrec 16440 15.4 3.880 4.856 9.711 10.556 qs_ot_get_derivative 99 11.5 0.001 0.001 9.281 9.376 init_scf_loop 11 6.9 0.012 0.095 8.631 8.642 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.582 7.723 prepare_preconditioner 11 7.9 0.000 0.000 6.712 6.731 make_preconditioner 11 8.9 0.000 0.001 6.712 6.731 sum_up_and_integrate 110 10.3 0.060 0.060 6.603 6.619 integrate_v_rspace 110 11.3 0.002 0.003 6.543 6.559 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.068 6.422 dbcsr_mm_accdrv_process 34862 16.1 4.632 5.618 5.685 6.154 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.999 6.008 calculate_rho_elec 110 8.6 0.058 0.059 5.999 6.007 init_scf_run 11 5.9 0.000 0.001 5.515 5.515 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.515 5.515 qs_ot_get_p 110 10.4 0.001 0.002 5.310 5.444 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.988 5.431 apply_single 110 13.6 0.000 0.000 4.988 5.430 make_m2s 4110 13.4 0.050 0.051 4.195 4.607 make_images 4110 14.4 0.391 0.506 4.080 4.493 ot_diis_step 99 11.5 0.010 0.011 4.437 4.438 density_rs2pw 110 9.6 0.004 0.005 3.133 4.331 multiply_cannon_sync_h2d 16440 15.4 3.696 4.204 3.696 4.204 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.404 4.047 rs_pw_transfer 902 11.9 0.011 0.012 2.735 3.934 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.711 3.714 pw_transfer 1331 11.6 0.066 0.074 3.688 3.698 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.581 3.593 grid_integrate_task_list 110 12.3 3.194 3.414 3.194 3.414 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.373 3.373 fft_wrap_pw1pw2_140 451 13.1 0.213 0.218 3.098 3.119 make_images_data 4110 15.4 0.043 0.048 2.472 2.970 wfi_extrapolate 11 7.9 0.001 0.001 2.963 2.963 fft3d_ps 1111 14.6 1.090 1.120 2.879 2.890 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.155 2.882 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.865 2.866 cp_fm_redistribute_end 48 14.0 1.788 2.841 1.791 2.841 cp_fm_diag_elpa_base 48 14.0 0.988 2.694 1.047 2.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.794 2.795 cp_fm_cholesky_invert 11 10.9 2.629 2.635 2.629 2.635 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.461 2.507 mp_waitany 17072 13.8 1.233 2.500 1.233 2.500 grid_collocate_task_list 110 9.6 2.085 2.470 2.085 2.470 calculate_first_density_matrix 1 7.0 0.000 0.002 2.454 2.455 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.882 2.425 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.292 2.347 potential_pw2rs 110 12.3 0.010 0.010 2.328 2.337 mp_irecv_dv 48980 15.7 0.812 2.302 0.812 2.302 mp_alltoall_d11v 2046 13.8 1.693 2.141 1.693 2.141 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.189 0.931 2.132 mp_sum_l 6514 12.8 1.505 2.112 1.505 2.112 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.971 1.985 dbcsr_complete_redistribute 325 12.2 0.327 0.364 1.466 1.933 cp_fm_upper_to_full 70 13.6 1.393 1.856 1.393 1.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.687 1.704 mp_allgather_i34 2055 14.4 0.529 1.674 0.529 1.674 cp_fm_cholesky_decompose 22 10.9 1.594 1.618 1.594 1.618 jit_kernel_multiply 8 16.4 0.668 1.547 0.668 1.547 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.490 1.504 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.500 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.024 1.483 rs_gather_matrices 110 12.3 0.234 0.260 0.947 1.383 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.242 1.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.614000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.636364, yerr=8.699767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 731.115520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.095 0.131 66.283 66.285 qs_mol_dyn_low 1 2.0 0.003 0.004 65.682 65.693 qs_forces 11 3.9 0.002 0.006 65.616 65.618 qs_energies 11 4.9 0.002 0.004 62.162 62.166 scf_env_do_scf 11 5.9 0.001 0.002 53.719 53.723 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.252 41.259 velocity_verlet 10 3.0 0.002 0.002 37.245 37.247 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.393 29.558 qs_scf_new_mos 99 7.5 0.001 0.001 26.090 26.193 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.090 26.192 ot_scf_mini 99 9.5 0.003 0.004 24.469 24.562 multiply_cannon 2055 13.4 0.245 0.260 22.544 23.586 multiply_cannon_loop 2055 14.4 0.886 0.902 20.985 21.471 ot_mini 99 10.5 0.001 0.001 13.948 14.056 multiply_cannon_multrec 24660 15.4 4.226 6.791 12.686 13.951 rebuild_ks_matrix 110 8.3 0.000 0.000 12.887 12.976 qs_ks_build_kohn_sham_matrix 110 9.3 0.142 0.537 12.887 12.976 init_scf_loop 11 6.9 0.001 0.003 12.424 12.442 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.439 11.518 prepare_preconditioner 11 7.9 0.000 0.000 10.007 10.021 make_preconditioner 11 8.9 0.000 0.001 10.007 10.021 qs_ot_get_derivative 99 11.5 0.001 0.001 9.811 9.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.201 9.679 dbcsr_mm_accdrv_process 52304 16.0 7.001 8.763 8.313 9.229 sum_up_and_integrate 110 10.3 0.068 0.071 6.576 6.591 integrate_v_rspace 110 11.3 0.003 0.004 6.507 6.522 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.219 6.228 calculate_rho_elec 110 8.6 0.077 0.081 6.219 6.227 mp_waitall_1 121746 16.5 4.131 6.209 4.131 6.209 qs_ot_get_p 110 10.4 0.001 0.002 5.607 5.768 make_m2s 4110 13.4 0.059 0.062 5.235 5.728 init_scf_run 11 5.9 0.000 0.001 5.646 5.646 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.645 5.646 make_images 4110 14.4 0.574 0.694 5.095 5.584 cp_fm_upper_to_full 70 13.8 3.307 4.710 3.307 4.710 ot_diis_step 99 11.5 0.011 0.011 4.102 4.103 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.016 4.091 apply_single 110 13.6 0.000 0.000 4.016 4.091 pw_transfer 1331 11.6 0.065 0.073 3.888 3.922 dbcsr_complete_redistribute 325 12.2 0.419 0.459 2.686 3.835 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.782 3.818 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.764 3.778 density_rs2pw 110 9.6 0.004 0.004 3.150 3.618 grid_integrate_task_list 110 12.3 3.270 3.432 3.270 3.432 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.287 3.339 multiply_cannon_sync_h2d 24660 15.4 3.191 3.299 3.191 3.299 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.294 3.294 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.153 3.288 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.239 3.278 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.347 3.132 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.229 3.132 fft3d_ps 1111 14.6 1.094 1.128 3.076 3.101 make_images_data 4110 15.4 0.046 0.051 2.595 3.086 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.981 3.045 calculate_dm_sparse 110 9.5 0.001 0.001 3.006 3.032 wfi_extrapolate 11 7.9 0.001 0.001 2.994 2.994 rs_pw_transfer 902 11.9 0.010 0.011 2.497 2.949 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.797 2.911 mp_alltoall_i22 605 13.7 1.664 2.853 1.664 2.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.786 2.787 cp_fm_cholesky_invert 11 10.9 2.751 2.760 2.751 2.760 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.741 2.744 cp_fm_redistribute_end 48 14.0 1.372 2.714 1.373 2.715 cp_fm_diag_elpa_base 48 14.0 1.261 2.579 1.339 2.687 grid_collocate_task_list 110 9.6 2.192 2.499 2.192 2.499 calculate_first_density_matrix 1 7.0 0.000 0.002 2.492 2.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.457 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.301 2.303 potential_pw2rs 110 12.3 0.013 0.013 2.253 2.262 jit_kernel_multiply 11 15.8 0.984 2.064 0.984 2.064 mp_alltoall_d11v 2046 13.8 1.725 2.002 1.725 2.002 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.688 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.605 1.716 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.693 1.706 cp_fm_cholesky_decompose 22 10.9 1.661 1.705 1.661 1.705 mp_waitany 13376 13.8 1.115 1.646 1.115 1.646 mp_allgather_i34 2055 14.4 0.661 1.642 0.661 1.642 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.623 1.638 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.833 1.576 mp_sum_l 6514 12.8 0.932 1.523 0.932 1.523 acc_transpose_blocks 24660 15.4 0.107 0.109 1.499 1.516 mp_irecv_dv 62702 16.1 0.734 1.500 0.734 1.500 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.249 1.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.285000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.545455, yerr=9.754846 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 853.389312E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 56.308 56.309 qs_mol_dyn_low 1 2.0 0.003 0.003 55.773 55.783 qs_forces 11 3.9 0.002 0.002 55.707 55.708 qs_energies 11 4.9 0.001 0.001 52.004 52.007 scf_env_do_scf 11 5.9 0.001 0.001 43.121 43.121 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.435 35.436 velocity_verlet 10 3.0 0.002 0.002 31.163 31.166 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.277 23.446 qs_scf_new_mos 99 7.5 0.001 0.001 20.305 20.366 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.305 20.366 multiply_cannon 2055 13.4 0.245 0.265 17.947 19.184 ot_scf_mini 99 9.5 0.002 0.002 19.077 19.106 multiply_cannon_loop 2055 14.4 0.320 0.331 16.616 17.032 rebuild_ks_matrix 110 8.3 0.000 0.000 11.969 12.007 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.969 12.006 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.628 10.662 ot_mini 99 10.5 0.001 0.001 10.413 10.443 multiply_cannon_multrec 8220 15.4 3.344 4.738 7.667 8.579 init_scf_loop 11 6.9 0.000 0.000 7.637 7.638 mp_waitall_1 103326 16.6 5.885 7.580 5.885 7.580 qs_ot_get_derivative 99 11.5 0.001 0.001 6.610 6.640 sum_up_and_integrate 110 10.3 0.079 0.081 6.552 6.569 integrate_v_rspace 110 11.3 0.003 0.003 6.472 6.491 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.339 6.353 calculate_rho_elec 110 8.6 0.115 0.116 6.339 6.353 prepare_preconditioner 11 7.9 0.000 0.000 5.971 5.976 make_preconditioner 11 8.9 0.000 0.000 5.971 5.976 init_scf_run 11 5.9 0.000 0.001 5.761 5.761 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.761 5.761 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.555 5.630 dbcsr_mm_accdrv_process 17442 15.9 2.981 4.180 4.193 5.275 qs_ot_get_p 110 10.4 0.001 0.001 4.790 4.835 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.059 4.507 make_m2s 4110 13.4 0.038 0.040 4.062 4.334 make_images 4110 14.4 0.636 0.691 3.933 4.201 pw_transfer 1331 11.6 0.066 0.071 4.103 4.113 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.996 4.010 ot_diis_step 99 11.5 0.012 0.012 3.783 3.783 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.717 3.764 apply_single 110 13.6 0.000 0.000 3.716 3.763 grid_integrate_task_list 110 12.3 3.338 3.540 3.338 3.540 density_rs2pw 110 9.6 0.004 0.004 3.111 3.487 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.483 3.487 fft_wrap_pw1pw2_140 451 13.1 0.216 0.223 3.457 3.476 fft3d_ps 1111 14.6 1.151 1.199 3.238 3.258 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.184 3.185 calculate_dm_sparse 110 9.5 0.001 0.001 3.007 3.049 multiply_cannon_sync_h2d 8220 15.4 2.922 3.027 2.922 3.027 cp_fm_cholesky_invert 11 10.9 2.923 2.927 2.923 2.927 calculate_first_density_matrix 1 7.0 0.000 0.000 2.901 2.902 wfi_extrapolate 11 7.9 0.001 0.001 2.762 2.762 make_images_data 4110 15.4 0.038 0.044 2.285 2.680 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.677 2.678 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.676 2.677 cp_fm_redistribute_end 48 14.0 0.680 2.649 0.684 2.649 cp_fm_diag_elpa_base 48 14.0 1.785 2.461 1.958 2.619 hybrid_alltoall_any 4261 16.3 0.200 0.861 2.224 2.580 rs_pw_transfer 902 11.9 0.010 0.011 2.173 2.573 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.553 2.555 grid_collocate_task_list 110 9.6 2.278 2.538 2.278 2.538 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.196 2.211 potential_pw2rs 110 12.3 0.015 0.016 2.114 2.119 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 1.998 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.979 1.990 mp_alltoall_d11v 2046 13.8 1.598 1.969 1.598 1.969 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.891 1.923 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.753 1.772 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.680 cp_fm_cholesky_decompose 22 10.9 1.629 1.647 1.629 1.647 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.511 1.632 mp_allgather_i34 2055 14.4 0.515 1.533 0.515 1.533 dbcsr_complete_redistribute 325 12.2 0.558 0.583 1.427 1.532 mp_waitany 9240 13.8 1.054 1.465 1.054 1.465 jit_kernel_multiply 8 15.7 0.906 1.455 0.906 1.455 acc_transpose_blocks 8220 15.4 0.036 0.037 1.355 1.372 qs_create_task_list 11 7.9 0.001 0.001 1.221 1.328 generate_qs_task_list 11 8.9 0.378 0.445 1.221 1.328 rs_gather_matrices 110 12.3 0.326 0.368 0.935 1.280 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.788 1.195 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.120 1.139 acc_transpose_blocks_kernels 8220 16.4 0.111 0.113 1.122 1.135 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.309000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=796.272727, yerr=17.077848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 1.398194E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 88.744 88.746 qs_mol_dyn_low 1 2.0 0.003 0.003 88.158 88.178 qs_forces 11 3.9 0.005 0.006 88.086 88.087 qs_energies 11 4.9 0.001 0.002 83.859 83.860 scf_env_do_scf 11 5.9 0.001 0.001 73.479 73.480 velocity_verlet 10 3.0 0.002 0.002 56.321 56.328 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.871 44.873 dbcsr_multiply_generic 2055 12.4 0.129 0.132 29.182 29.233 init_scf_loop 11 6.9 0.000 0.000 28.534 28.538 prepare_preconditioner 11 7.9 0.000 0.000 26.486 26.492 make_preconditioner 11 8.9 0.000 0.000 26.486 26.492 qs_scf_new_mos 99 7.5 0.001 0.001 26.390 26.420 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.389 26.419 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.624 25.953 ot_scf_mini 99 9.5 0.002 0.002 24.631 24.657 multiply_cannon 2055 13.4 0.341 0.350 22.050 22.767 multiply_cannon_loop 2055 14.4 0.341 0.344 20.251 20.657 cp_fm_upper_to_full 70 14.2 12.860 18.430 12.860 18.430 rebuild_ks_matrix 110 8.3 0.001 0.001 14.293 14.338 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.293 14.337 ot_mini 99 10.5 0.001 0.001 13.808 13.830 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.947 12.986 dbcsr_complete_redistribute 325 12.2 1.022 1.046 7.430 10.663 multiply_cannon_multrec 8220 15.4 4.346 4.537 9.566 9.703 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.407 9.636 qs_ot_get_derivative 99 11.5 0.001 0.001 9.245 9.270 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.846 9.047 mp_waitall_1 84994 16.7 7.757 8.738 7.757 8.738 mp_alltoall_i22 605 13.7 5.464 8.675 5.464 8.675 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.812 7.850 calculate_rho_elec 110 8.6 0.227 0.227 7.811 7.849 sum_up_and_integrate 110 10.3 0.150 0.152 7.413 7.429 integrate_v_rspace 110 11.3 0.003 0.004 7.263 7.277 init_scf_run 11 5.9 0.000 0.001 6.091 6.091 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.090 6.090 make_m2s 4110 13.4 0.043 0.043 5.371 5.888 make_images 4110 14.4 0.876 0.918 5.182 5.699 qs_ot_get_p 110 10.4 0.001 0.001 5.602 5.628 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.869 5.413 pw_transfer 1331 11.6 0.075 0.075 5.328 5.335 dbcsr_mm_accdrv_process 11614 15.7 3.335 3.635 5.077 5.325 cp_fm_cholesky_invert 11 10.9 5.278 5.283 5.278 5.283 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.212 5.219 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.685 5.150 apply_single 110 13.6 0.000 0.000 4.685 5.150 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.535 4.546 ot_diis_step 99 11.5 0.015 0.015 4.539 4.540 fft3d_ps 1111 14.6 1.292 1.300 4.408 4.417 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.066 4.075 multiply_cannon_sync_h2d 8220 15.4 3.947 3.952 3.947 3.952 density_rs2pw 110 9.6 0.004 0.004 3.844 3.876 hybrid_alltoall_any 4261 16.3 0.257 0.557 2.987 3.765 grid_integrate_task_list 110 12.3 3.666 3.718 3.666 3.718 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.696 3.697 make_images_data 4110 15.4 0.042 0.045 2.998 3.652 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.623 3.624 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.159 3.611 calculate_dm_sparse 110 9.5 0.001 0.001 3.487 3.507 wfi_extrapolate 11 7.9 0.001 0.001 3.301 3.301 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.020 3.020 cp_fm_diag_elpa_base 48 14.0 2.483 2.677 3.018 3.018 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.887 2.890 calculate_first_density_matrix 1 7.0 0.000 0.000 2.666 2.667 grid_collocate_task_list 110 9.6 2.632 2.658 2.632 2.658 potential_pw2rs 110 12.3 0.021 0.022 2.527 2.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.395 2.410 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.375 2.389 rs_pw_transfer 902 11.9 0.011 0.011 2.292 2.336 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.220 2.286 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.208 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.022 2.038 mp_alltoall_d11v 2046 13.8 1.935 2.019 1.935 2.019 cp_fm_cholesky_decompose 22 10.9 1.971 1.988 1.971 1.988 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.951 1.956 qs_create_task_list 11 7.9 0.000 0.001 1.899 1.947 generate_qs_task_list 11 8.9 0.735 0.790 1.898 1.947 jit_kernel_multiply 10 15.2 1.543 1.805 1.543 1.805 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.755 1.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.746000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1240.909091, yerr=79.399398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 631.140352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4081 56811. MP_Allreduce 11243 786. MP_Sync 168 MP_Alltoall 2210 2520183. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.051 203.476 203.478 qs_mol_dyn_low 1 2.0 0.007 0.032 203.000 203.013 qs_forces 11 3.9 0.004 0.015 202.850 202.852 qs_energies 11 4.9 0.001 0.003 197.158 197.171 scf_env_do_scf 11 5.9 0.001 0.001 180.358 180.362 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 159.564 159.566 dbcsr_multiply_generic 2485 12.5 0.176 0.180 123.100 123.841 velocity_verlet 10 3.0 0.001 0.002 121.585 121.586 qs_scf_new_mos 116 7.6 0.001 0.001 119.507 119.816 qs_scf_loop_do_ot 116 8.6 0.001 0.001 119.506 119.816 ot_scf_mini 116 9.6 0.003 0.003 113.013 113.330 multiply_cannon 2485 13.5 0.233 0.242 100.203 102.452 multiply_cannon_loop 2485 14.5 2.095 2.152 98.051 100.373 ot_mini 116 10.6 0.001 0.001 64.842 65.143 multiply_cannon_multrec 59640 15.5 32.870 35.072 41.137 43.579 qs_ot_get_derivative 116 11.6 0.001 0.001 40.363 40.646 rebuild_ks_matrix 127 8.3 0.001 0.001 33.672 34.002 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 33.672 34.002 mp_waitall_1 264810 16.5 28.119 31.433 28.119 31.433 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.235 30.536 multiply_cannon_sync_h2d 59640 15.5 27.007 28.574 27.007 28.574 qs_ot_get_p 127 10.4 0.001 0.001 26.576 26.833 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 23.981 24.976 apply_single 127 13.6 0.001 0.001 23.980 24.976 ot_diis_step 116 11.6 0.007 0.008 24.129 24.130 init_scf_loop 11 6.9 0.000 0.000 20.717 20.718 qs_ot_p2m_diag 82 11.4 0.076 0.091 20.042 20.164 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.659 18.888 multiply_cannon_metrocomm3 59640 15.5 0.115 0.120 15.422 18.213 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.451 17.452 prepare_preconditioner 11 7.9 0.000 0.000 16.064 16.109 make_preconditioner 11 8.9 0.000 0.000 16.064 16.109 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.269 15.462 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.657 14.663 cp_fm_redistribute_end 82 14.4 11.587 14.588 11.600 14.594 cp_fm_diag_elpa_base 82 14.4 2.947 14.222 2.979 14.336 sum_up_and_integrate 127 10.3 0.089 0.107 14.297 14.310 integrate_v_rspace 127 11.3 0.003 0.004 14.208 14.223 make_m2s 4970 13.5 0.104 0.114 13.623 14.024 qs_rho_update_rho_low 127 7.7 0.001 0.002 13.810 13.901 calculate_rho_elec 127 8.7 0.045 0.064 13.809 13.901 make_images 4970 14.5 0.395 0.413 13.442 13.850 init_scf_run 11 5.9 0.000 0.001 12.683 12.683 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.683 12.683 density_rs2pw 127 9.7 0.006 0.007 7.179 10.359 mp_sum_l 7804 13.0 8.030 9.605 8.030 9.605 cp_fm_cholesky_invert 11 10.9 9.211 9.219 9.211 9.219 wfi_extrapolate 11 7.9 0.001 0.001 9.090 9.090 rs_pw_transfer 1038 11.9 0.016 0.018 5.767 8.943 multiply_cannon_metrocomm1 59640 15.5 0.089 0.094 6.204 8.905 calculate_dm_sparse 127 9.5 0.001 0.001 8.287 8.430 dbcsr_mm_accdrv_process 123452 16.2 3.267 3.432 7.834 8.372 pw_transfer 1535 11.6 0.076 0.104 8.012 8.245 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.019 8.137 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 7.808 8.047 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.652 7.777 make_images_data 4970 15.5 0.065 0.071 6.716 7.646 grid_integrate_task_list 127 12.3 7.007 7.526 7.007 7.526 hybrid_alltoall_any 5155 16.4 0.288 2.246 5.883 7.187 fft_wrap_pw1pw2_140 519 13.2 0.442 0.510 6.808 7.095 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 7.035 7.042 fft3d_ps 1281 14.7 2.086 2.860 6.604 6.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.193 6.332 mp_waitany 15900 13.9 2.701 5.985 2.701 5.985 grid_collocate_task_list 127 9.7 4.516 5.731 4.516 5.731 mp_alltoall_d11v 2401 14.1 4.272 5.479 4.272 5.479 rs_pw_transfer_RS2PW_140 138 11.5 0.276 0.292 2.115 5.317 potential_pw2rs 127 12.3 0.009 0.011 4.735 4.760 cp_fm_cholesky_decompose 22 10.9 4.631 4.648 4.631 4.648 mp_sum_d 4442 12.1 3.594 4.302 3.594 4.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.478000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=7.968948 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 834.342912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5618989. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 190.552 190.553 qs_mol_dyn_low 1 2.0 0.003 0.003 190.006 190.020 qs_forces 11 3.9 0.003 0.003 189.921 189.923 qs_energies 11 4.9 0.001 0.002 183.093 183.104 scf_env_do_scf 11 5.9 0.001 0.001 166.952 166.961 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.165 134.167 velocity_verlet 10 3.0 0.005 0.006 119.856 119.857 dbcsr_multiply_generic 2507 12.6 0.186 0.192 96.667 97.965 qs_scf_new_mos 117 7.6 0.001 0.001 93.909 94.457 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.908 94.456 ot_scf_mini 117 9.6 0.004 0.004 89.110 89.704 multiply_cannon 2507 13.6 0.477 0.528 76.922 80.989 multiply_cannon_loop 2507 14.6 1.256 1.311 74.024 76.648 ot_mini 117 10.6 0.001 0.001 49.469 50.002 mp_waitall_1 214728 16.6 24.315 37.190 24.315 37.190 multiply_cannon_multrec 30084 15.6 22.000 26.500 31.584 36.550 rebuild_ks_matrix 128 8.3 0.001 0.001 33.397 33.927 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 33.396 33.926 init_scf_loop 11 6.9 0.000 0.002 32.700 32.702 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.052 30.538 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.743 28.692 prepare_preconditioner 11 7.9 0.000 0.000 28.306 28.393 make_preconditioner 11 8.9 0.000 0.000 28.306 28.393 qs_ot_get_derivative 117 11.6 0.001 0.002 27.561 28.160 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.978 27.491 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.978 23.075 apply_single 128 13.6 0.001 0.001 21.978 23.074 qs_ot_get_p 128 10.4 0.001 0.001 21.565 22.382 multiply_cannon_sync_h2d 30084 15.6 19.286 21.975 19.286 21.975 ot_diis_step 117 11.6 0.014 0.015 21.732 21.734 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.809 16.845 cp_fm_cholesky_invert 11 10.9 16.522 16.534 16.522 16.534 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.673 15.674 make_m2s 5014 13.6 0.090 0.097 14.081 15.209 make_images 5014 14.6 1.146 1.327 13.870 14.998 sum_up_and_integrate 128 10.3 0.116 0.133 14.871 14.901 integrate_v_rspace 128 11.3 0.003 0.003 14.755 14.791 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.194 14.220 calculate_rho_elec 128 8.7 0.088 0.105 14.194 14.219 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.583 12.593 cp_fm_redistribute_end 83 14.4 7.348 12.519 7.359 12.522 cp_fm_diag_elpa_base 83 14.4 4.928 12.021 5.144 12.381 init_scf_run 11 5.9 0.000 0.001 11.381 11.382 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.380 11.382 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.823 11.213 multiply_cannon_metrocomm4 27577 15.6 0.095 0.110 3.724 10.575 mp_irecv_dv 69486 16.3 3.531 10.186 3.531 10.186 density_rs2pw 128 9.7 0.006 0.007 7.559 9.988 make_images_data 5014 15.6 0.063 0.072 8.521 9.929 dbcsr_mm_accdrv_process 62242 16.2 4.457 5.185 9.045 9.646 hybrid_alltoall_any 5200 16.5 0.342 1.515 7.280 9.328 pw_transfer 1547 11.6 0.087 0.109 9.091 9.172 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.865 8.953 wfi_extrapolate 11 7.9 0.001 0.001 8.348 8.349 rs_pw_transfer 1046 11.9 0.014 0.017 5.684 8.120 fft_wrap_pw1pw2_140 523 13.2 0.474 0.533 7.839 7.936 grid_integrate_task_list 128 12.3 7.179 7.547 7.179 7.547 fft3d_ps 1291 14.7 2.756 2.945 7.232 7.275 cp_fm_cholesky_decompose 22 10.9 6.845 6.937 6.845 6.937 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.156 6.887 calculate_dm_sparse 128 9.5 0.001 0.001 6.448 6.593 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.045 6.055 grid_collocate_task_list 128 9.7 4.705 5.882 4.705 5.882 mp_sum_l 7870 13.0 3.876 5.776 3.876 5.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.229 5.391 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.320 5.390 potential_pw2rs 128 12.3 0.014 0.016 5.099 5.124 mp_waitany 11748 13.9 2.469 4.938 2.469 4.938 mp_alltoall_d11v 2415 14.1 4.130 4.883 4.130 4.883 rs_pw_transfer_RS2PW_140 139 11.5 0.350 0.384 2.074 4.494 mp_allgather_i34 2507 14.6 1.468 4.386 1.468 4.386 mp_sum_d 4457 12.1 2.630 4.071 2.630 4.071 dbcsr_complete_redistribute 395 12.7 0.773 0.885 3.132 3.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.553000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.454545, yerr=3.285329 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 933.941248E+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 931530938576 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_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.151 0.758 178.659 178.661 qs_mol_dyn_low 1 2.0 0.003 0.004 177.467 177.479 qs_forces 11 3.9 0.003 0.008 177.347 177.353 qs_energies 11 4.9 0.002 0.006 170.622 170.631 scf_env_do_scf 11 5.9 0.001 0.002 154.654 154.655 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.650 119.651 velocity_verlet 10 3.0 0.002 0.002 113.352 113.354 dbcsr_multiply_generic 2507 12.6 0.183 0.189 82.383 83.461 qs_scf_new_mos 117 7.6 0.001 0.001 81.194 81.522 qs_scf_loop_do_ot 117 8.6 0.002 0.008 81.193 81.521 ot_scf_mini 117 9.6 0.004 0.005 77.082 77.467 multiply_cannon 2507 13.6 0.500 0.529 62.556 66.593 multiply_cannon_loop 2507 14.6 0.855 0.881 59.316 61.865 ot_mini 117 10.6 0.001 0.001 42.872 43.270 mp_waitall_1 170520 16.6 24.884 35.012 24.884 35.012 init_scf_loop 11 6.9 0.001 0.004 34.908 34.910 rebuild_ks_matrix 128 8.3 0.001 0.001 31.032 31.440 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 31.031 31.440 prepare_preconditioner 11 7.9 0.000 0.001 30.830 30.873 make_preconditioner 11 8.9 0.000 0.003 30.830 30.873 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.500 29.918 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.896 28.269 multiply_cannon_multrec 20056 15.6 13.615 16.386 22.635 25.508 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 14.978 24.756 qs_ot_get_derivative 117 11.6 0.001 0.002 23.203 23.598 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.761 20.680 apply_single 128 13.6 0.001 0.001 19.761 20.679 qs_ot_get_p 128 10.4 0.001 0.002 19.521 19.959 ot_diis_step 117 11.6 0.018 0.019 19.570 19.570 multiply_cannon_sync_h2d 20056 15.6 14.293 16.093 14.293 16.093 make_m2s 5014 13.6 0.081 0.085 14.863 15.898 make_images 5014 14.6 1.179 1.264 14.629 15.664 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.309 15.317 sum_up_and_integrate 128 10.3 0.135 0.146 14.779 14.817 integrate_v_rspace 128 11.3 0.004 0.005 14.644 14.676 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.417 14.462 calculate_rho_elec 128 8.7 0.132 0.147 14.417 14.462 cp_fm_cholesky_invert 11 10.9 14.329 14.338 14.329 14.338 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.311 14.312 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.151 11.152 cp_fm_redistribute_end 83 14.4 4.220 11.091 4.236 11.093 cp_fm_diag_elpa_base 83 14.4 6.430 10.495 6.838 10.966 init_scf_run 11 5.9 0.000 0.001 10.959 10.960 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.959 10.959 make_images_data 5014 15.6 0.061 0.069 9.225 10.790 hybrid_alltoall_any 5200 16.5 0.433 1.996 7.970 9.982 pw_transfer 1547 11.6 0.086 0.107 9.145 9.269 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.875 9.154 density_rs2pw 128 9.7 0.006 0.006 7.286 9.122 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.403 9.105 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.920 9.047 mp_irecv_dv 50230 16.2 3.282 8.858 3.282 8.858 dbcsr_mm_accdrv_process 41502 16.2 4.504 5.629 8.469 8.744 fft_wrap_pw1pw2_140 523 13.2 0.477 0.526 7.908 8.042 grid_integrate_task_list 128 12.3 7.321 7.725 7.321 7.725 cp_fm_upper_to_full 105 14.5 5.808 7.543 5.808 7.543 wfi_extrapolate 11 7.9 0.001 0.001 7.427 7.427 cp_fm_cholesky_decompose 22 10.9 7.340 7.374 7.340 7.374 fft3d_ps 1291 14.7 2.664 2.890 7.185 7.277 rs_pw_transfer 1046 11.9 0.014 0.015 5.162 7.018 dbcsr_complete_redistribute 395 12.7 1.175 1.202 4.524 6.307 calculate_dm_sparse 128 9.5 0.001 0.001 5.862 5.940 grid_collocate_task_list 128 9.7 4.904 5.781 4.904 5.781 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.766 5.774 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.769 5.435 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.384 5.160 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.883 4.998 potential_pw2rs 128 12.3 0.020 0.023 4.892 4.917 mp_alltoall_d11v 2415 14.1 4.211 4.882 4.211 4.882 mp_sum_l 7870 13.0 3.148 4.670 3.148 4.670 mp_allgather_i34 2507 14.6 1.758 4.666 1.758 4.666 mp_waitany 11748 13.9 2.313 4.222 2.313 4.222 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.015 4.050 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.310 4.028 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.768 3.769 rs_pw_transfer_RS2PW_140 139 11.5 0.329 0.349 1.905 3.750 mp_alltoall_i22 716 14.1 1.928 3.742 1.928 3.742 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.647 3.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.661000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=884.818182, yerr=11.495597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.119101E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.077 187.808 187.827 qs_mol_dyn_low 1 2.0 0.003 0.005 187.099 187.113 qs_forces 11 3.9 0.004 0.010 186.985 186.990 qs_energies 11 4.9 0.005 0.027 179.839 179.847 scf_env_do_scf 11 5.9 0.022 0.153 163.099 163.109 velocity_verlet 10 3.0 0.003 0.004 123.757 123.760 scf_env_do_scf_inner_loop 117 6.6 0.007 0.019 116.395 116.396 dbcsr_multiply_generic 2507 12.6 0.190 0.195 79.096 79.842 qs_scf_new_mos 117 7.6 0.001 0.001 79.127 79.422 qs_scf_loop_do_ot 117 8.6 0.003 0.008 79.126 79.421 ot_scf_mini 117 9.6 0.004 0.004 74.664 74.966 multiply_cannon 2507 13.6 0.559 0.596 54.896 57.516 multiply_cannon_loop 2507 14.6 1.183 1.210 51.295 53.094 init_scf_loop 11 6.9 0.001 0.006 46.432 46.438 ot_mini 117 10.6 0.001 0.001 42.308 42.596 prepare_preconditioner 11 7.9 0.000 0.000 42.140 42.161 make_preconditioner 11 8.9 0.000 0.001 42.140 42.161 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.775 40.778 multiply_cannon_multrec 30084 15.6 14.293 19.746 26.567 31.568 rebuild_ks_matrix 128 8.3 0.001 0.001 29.924 30.219 qs_ks_build_kohn_sham_matrix 128 9.3 0.046 0.134 29.923 30.218 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.000 27.251 mp_waitall_1 147882 16.7 16.726 26.927 16.726 26.927 qs_ot_get_derivative 117 11.6 0.001 0.002 22.740 23.038 make_m2s 5014 13.6 0.096 0.099 19.908 20.827 make_images 5014 14.6 1.935 2.246 19.601 20.518 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.991 19.600 apply_single 128 13.6 0.001 0.001 18.990 19.599 ot_diis_step 117 11.6 0.017 0.018 19.441 19.442 qs_ot_get_p 128 10.4 0.001 0.001 18.673 19.000 cp_fm_upper_to_full 105 14.7 11.025 16.350 11.025 16.350 cp_fm_cholesky_invert 11 10.9 16.115 16.124 16.115 16.124 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.233 14.991 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.700 14.751 sum_up_and_integrate 128 10.3 0.141 0.152 14.654 14.681 integrate_v_rspace 128 11.3 0.009 0.035 14.513 14.543 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.495 14.530 calculate_rho_elec 128 8.7 0.175 0.191 14.495 14.530 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.436 13.437 multiply_cannon_sync_h2d 30084 15.6 11.709 12.987 11.709 12.987 dbcsr_complete_redistribute 395 12.7 1.509 1.622 8.842 12.557 dbcsr_mm_accdrv_process 62264 16.2 7.718 8.804 11.848 12.392 make_images_data 5014 15.6 0.064 0.070 10.673 12.286 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.504 11.241 hybrid_alltoall_any 5200 16.5 0.527 2.224 9.489 11.188 init_scf_run 11 5.9 0.000 0.001 10.842 10.844 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.842 10.844 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.353 10.355 cp_fm_redistribute_end 83 14.4 1.787 10.279 1.800 10.285 cp_fm_diag_elpa_base 83 14.4 7.880 9.685 8.452 10.165 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.341 9.977 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.193 9.406 mp_alltoall_i22 716 14.1 5.683 9.400 5.683 9.400 pw_transfer 1547 11.6 0.088 0.107 9.168 9.246 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.940 9.028 density_rs2pw 128 9.7 0.006 0.006 7.012 8.564 fft_wrap_pw1pw2_140 523 13.2 0.485 0.493 7.936 8.032 grid_integrate_task_list 128 12.3 7.493 8.018 7.493 8.018 cp_fm_cholesky_decompose 22 10.9 7.527 7.622 7.527 7.622 wfi_extrapolate 11 7.9 0.001 0.001 7.409 7.410 fft3d_ps 1291 14.7 2.778 2.884 7.163 7.245 multiply_cannon_metrocomm4 25070 15.6 0.076 0.085 2.754 7.177 mp_irecv_dv 76098 16.2 2.610 6.909 2.610 6.909 rs_pw_transfer 1046 11.9 0.013 0.015 4.670 6.295 calculate_dm_sparse 128 9.5 0.001 0.001 6.098 6.185 grid_collocate_task_list 128 9.7 5.101 5.807 5.101 5.807 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.458 mp_alltoall_d11v 2415 14.1 4.568 5.260 4.568 5.260 potential_pw2rs 128 12.3 0.023 0.023 4.559 4.576 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.459 4.532 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.394 4.498 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.463 4.464 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.155 4.220 mp_waitany 14952 13.9 2.276 3.900 2.276 3.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.827000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1057.818182, yerr=12.981232 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.496322E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.048 174.868 174.874 qs_mol_dyn_low 1 2.0 0.003 0.003 174.482 174.496 qs_forces 11 3.9 0.003 0.003 174.372 174.376 qs_energies 11 4.9 0.002 0.002 166.822 166.831 scf_env_do_scf 11 5.9 0.001 0.001 148.918 148.920 velocity_verlet 10 3.0 0.006 0.006 113.567 113.571 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.249 113.250 dbcsr_multiply_generic 2507 12.6 0.183 0.188 74.845 75.248 qs_scf_new_mos 117 7.6 0.001 0.001 74.585 74.724 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.585 74.724 ot_scf_mini 117 9.6 0.004 0.004 70.189 70.246 multiply_cannon 2507 13.6 0.583 0.614 55.155 59.816 multiply_cannon_loop 2507 14.6 0.443 0.456 50.328 51.409 ot_mini 117 10.6 0.001 0.001 39.970 40.026 init_scf_loop 11 6.9 0.000 0.000 35.526 35.528 mp_waitall_1 125778 16.7 25.303 32.895 25.303 32.895 prepare_preconditioner 11 7.9 0.000 0.000 31.518 31.551 make_preconditioner 11 8.9 0.000 0.000 31.518 31.551 rebuild_ks_matrix 128 8.3 0.001 0.001 29.968 30.108 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.967 30.108 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.373 29.648 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.227 27.352 multiply_cannon_multrec 10028 15.6 10.424 14.150 18.482 21.738 qs_ot_get_derivative 117 11.6 0.001 0.002 20.136 20.197 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.537 19.780 apply_single 128 13.6 0.001 0.001 19.536 19.779 ot_diis_step 117 11.6 0.019 0.020 19.768 19.768 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.267 19.683 make_m2s 5014 13.6 0.065 0.071 15.920 18.411 make_images 5014 14.6 2.282 2.599 15.614 18.107 cp_fm_cholesky_invert 11 10.9 17.994 18.000 17.994 18.000 qs_ot_get_p 128 10.4 0.001 0.001 17.363 17.455 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.386 15.436 calculate_rho_elec 128 8.7 0.259 0.268 15.385 15.435 sum_up_and_integrate 128 10.3 0.181 0.191 15.067 15.112 integrate_v_rspace 128 11.3 0.004 0.004 14.885 14.939 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.617 13.636 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.496 12.499 multiply_cannon_sync_h2d 10028 15.6 11.619 12.382 11.619 12.382 make_images_data 5014 15.6 0.052 0.060 9.719 12.148 hybrid_alltoall_any 5200 16.5 0.819 3.638 9.555 11.859 init_scf_run 11 5.9 0.000 0.001 11.153 11.154 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.153 11.153 pw_transfer 1547 11.6 0.086 0.095 9.846 9.891 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.623 9.673 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.477 9.487 cp_fm_diag_elpa_base 83 14.4 9.236 9.317 9.471 9.481 fft_wrap_pw1pw2_140 523 13.2 0.498 0.525 8.496 8.551 dbcsr_mm_accdrv_process 20762 16.1 2.757 3.923 7.687 8.473 grid_integrate_task_list 128 12.3 7.724 8.174 7.724 8.174 mp_allgather_i34 2507 14.6 3.028 7.972 3.028 7.972 cp_fm_cholesky_decompose 22 10.9 7.857 7.950 7.857 7.950 density_rs2pw 128 9.7 0.005 0.006 6.996 7.889 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.800 7.843 fft3d_ps 1291 14.7 2.719 2.812 7.780 7.802 wfi_extrapolate 11 7.9 0.001 0.001 7.375 7.375 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.085 6.759 calculate_dm_sparse 128 9.5 0.001 0.001 6.040 6.142 grid_collocate_task_list 128 9.7 5.392 6.081 5.392 6.081 mp_alltoall_d11v 2415 14.1 4.893 5.670 4.893 5.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.629 5.639 dbcsr_complete_redistribute 395 12.7 2.138 2.193 5.115 5.530 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.293 5.293 rs_pw_transfer 1046 11.9 0.013 0.013 4.074 5.001 potential_pw2rs 128 12.3 0.027 0.027 4.588 4.611 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.572 4.594 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.964 4.127 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.848 4.067 mp_irecv_dv 28860 15.9 1.812 3.997 1.812 3.997 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.570 3.853 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.664 3.719 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.369 3.692 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.646 3.661 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.474 3.567 calculate_first_density_matrix 1 7.0 0.000 0.000 3.565 3.567 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.874000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1402.454545, yerr=40.839739 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 2.939912E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.053 299.510 299.513 qs_mol_dyn_low 1 2.0 0.003 0.003 298.977 298.990 qs_forces 11 3.9 0.003 0.003 298.838 298.840 qs_energies 11 4.9 0.005 0.005 289.448 289.462 scf_env_do_scf 11 5.9 0.001 0.001 266.528 266.542 velocity_verlet 10 3.0 0.005 0.006 215.698 215.711 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 142.020 142.021 init_scf_loop 11 6.9 0.000 0.000 124.256 124.259 prepare_preconditioner 11 7.9 0.000 0.000 119.130 119.159 make_preconditioner 11 8.9 0.000 0.000 119.130 119.159 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.036 116.299 qs_scf_new_mos 118 7.6 0.001 0.001 90.397 90.512 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.396 90.511 ot_scf_mini 118 9.6 0.004 0.004 85.619 85.657 dbcsr_multiply_generic 2529 12.6 0.216 0.222 83.715 84.251 cp_fm_upper_to_full 106 14.8 52.939 76.033 52.939 76.033 multiply_cannon 2529 13.6 0.718 0.807 59.156 59.646 multiply_cannon_loop 2529 14.6 0.480 0.515 55.607 56.874 ot_mini 118 10.6 0.001 0.001 44.244 44.281 dbcsr_complete_redistribute 397 12.7 3.972 4.025 29.654 42.741 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.322 39.382 rebuild_ks_matrix 129 8.3 0.001 0.001 38.927 38.959 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.927 38.958 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.049 36.981 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.909 35.943 mp_alltoall_i22 720 14.1 21.843 35.067 21.843 35.067 cp_fm_cholesky_invert 11 10.9 33.030 33.036 33.030 33.036 mp_waitall_1 104580 16.8 28.075 31.908 28.075 31.908 qs_ot_get_p 129 10.4 0.001 0.001 25.825 25.830 qs_ot_get_derivative 118 11.6 0.002 0.002 24.239 24.278 qs_ot_p2m_diag 84 11.4 0.889 0.893 21.827 21.854 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.908 20.923 calculate_rho_elec 129 8.7 0.486 0.487 20.907 20.922 make_m2s 5058 13.6 0.077 0.080 19.884 20.693 make_images 5058 14.6 3.801 3.966 19.401 20.211 sum_up_and_integrate 129 10.3 0.323 0.325 20.055 20.131 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.103 20.104 ot_diis_step 118 11.6 0.023 0.031 19.979 19.980 multiply_cannon_metrocomm3 10116 15.6 0.024 0.024 18.803 19.967 integrate_v_rspace 129 11.3 0.004 0.004 19.732 19.809 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.228 19.432 apply_single 129 13.6 0.001 0.001 19.228 19.432 multiply_cannon_multrec 10116 15.6 10.523 12.295 18.479 18.548 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.882 16.883 cp_fm_diag_elpa_base 84 14.4 12.526 14.135 16.878 16.878 multiply_cannon_sync_h2d 10116 15.6 15.776 15.788 15.776 15.788 init_scf_run 11 5.9 0.000 0.001 12.655 12.656 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.655 12.656 hybrid_alltoall_any 5245 16.5 1.310 3.061 10.760 12.556 pw_transfer 1559 11.6 0.093 0.094 12.357 12.362 make_images_data 5058 15.6 0.061 0.066 10.554 12.238 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.120 12.124 fft_wrap_pw1pw2_140 527 13.2 0.546 0.547 10.773 10.782 fft3d_ps 1301 14.7 2.759 2.769 10.145 10.150 dbcsr_mm_accdrv_process 20934 16.1 3.828 5.770 7.713 9.548 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.474 9.508 mp_alltoall_d11v 2429 14.1 8.308 9.256 8.308 9.256 wfi_extrapolate 11 7.9 0.001 0.001 9.073 9.073 grid_integrate_task_list 129 12.3 8.656 8.820 8.656 8.820 cp_fm_cholesky_decompose 22 10.9 8.769 8.790 8.769 8.790 density_rs2pw 129 9.7 0.005 0.005 8.361 8.508 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.304 8.305 calculate_dm_sparse 129 9.5 0.001 0.001 7.117 7.215 grid_collocate_task_list 129 9.7 6.415 6.449 6.415 6.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.299 6.370 rs_scatter_matrices 140 9.7 3.687 4.668 6.098 6.344 copy_dbcsr_to_fm 187 11.8 0.004 0.005 5.989 6.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.513000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2604.636364, yerr=179.814183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 1.261613E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 85.211 85.211 qs_energies 1 2.0 0.000 0.000 84.768 84.777 ls_scf 1 3.0 0.000 0.000 83.851 83.861 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.933 73.161 multiply_cannon 111 7.7 0.017 0.020 56.139 57.428 multiply_cannon_loop 111 8.7 0.209 0.221 52.722 53.976 ls_scf_main 1 4.0 0.000 0.000 52.213 52.213 density_matrix_trs4 2 5.0 0.002 0.003 46.689 46.777 ls_scf_init_scf 1 4.0 0.000 0.000 28.611 28.612 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.487 27.542 mp_waitall_1 11031 10.9 22.592 25.949 22.592 25.949 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.130 25.152 multiply_cannon_multrec 2664 9.7 8.186 8.944 15.479 17.154 multiply_cannon_sync_h2d 2664 9.7 13.596 15.605 13.596 15.605 make_m2s 222 7.7 0.008 0.010 12.934 13.477 make_images 222 8.7 0.099 0.108 12.912 13.457 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.822 12.585 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.565 8.692 make_images_data 222 9.7 0.004 0.005 7.529 8.057 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.641 6.911 7.859 hybrid_alltoall_any 227 10.6 0.214 1.821 6.478 7.829 dbcsr_mm_accdrv_process_sort 4760 11.4 6.199 7.068 6.199 7.068 calculate_norms 4752 9.8 5.501 6.111 5.501 6.111 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.267 5.359 mp_sum_l 807 5.4 3.321 4.281 3.321 4.281 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.064 3.595 mp_irecv_dv 6231 10.9 2.047 3.569 2.047 3.569 make_images_sizes 222 9.7 0.000 0.000 0.707 3.450 mp_alltoall_i44 222 10.7 0.707 3.450 0.707 3.450 arnoldi_extremal 4 6.8 0.000 0.000 3.190 3.223 arnoldi_normal_ev 4 7.8 0.001 0.002 3.190 3.223 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.368 3.182 build_subspace 16 8.4 0.009 0.012 3.097 3.099 ls_scf_post 1 4.0 0.000 0.000 3.028 3.037 ls_scf_store_result 1 5.0 0.000 0.000 2.841 2.885 dbcsr_special_finalize 555 9.7 0.005 0.006 2.345 2.780 dbcsr_merge_single_wm 555 10.7 0.452 0.587 2.337 2.772 make_images_pack 222 9.7 2.208 2.631 2.210 2.633 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.309 2.549 dbcsr_sort_data 658 11.4 2.143 2.522 2.143 2.522 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.453 2.065 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.298 2.393 buffer_matrices_ensure_size 222 8.7 1.752 2.090 1.752 2.090 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.793 1.795 rebuild_ks_matrix 3 7.3 0.000 0.000 1.784 1.785 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.784 1.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.211000, 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/6b3d50053968d9441899c157df378c42aede3ea5_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 2.103534E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.154 92.315 92.323 qs_energies 1 2.0 0.001 0.006 91.302 91.316 ls_scf 1 3.0 0.002 0.012 89.837 89.847 dbcsr_multiply_generic 111 6.7 0.016 0.023 75.256 75.718 multiply_cannon 111 7.7 0.037 0.115 53.252 56.858 ls_scf_main 1 4.0 0.000 0.001 54.942 54.969 multiply_cannon_loop 111 8.7 0.116 0.122 49.892 52.650 density_matrix_trs4 2 5.0 0.023 0.222 48.965 49.181 ls_scf_init_scf 1 4.0 0.000 0.001 31.337 31.478 mp_waitall_1 9105 10.9 21.140 30.034 21.140 30.034 ls_scf_init_matrix_S 1 5.0 0.005 0.040 29.830 29.925 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.361 27.377 multiply_cannon_multrec 1332 9.7 13.157 16.055 22.365 26.404 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.821 20.846 make_m2s 222 7.7 0.006 0.007 15.358 16.089 make_images 222 8.7 1.565 1.918 15.329 16.062 dbcsr_mm_accdrv_process 4041 10.4 0.284 0.441 8.805 10.320 dbcsr_mm_accdrv_process_sort 4041 11.4 8.398 9.896 8.398 9.896 make_images_data 222 9.7 0.005 0.013 8.803 9.838 hybrid_alltoall_any 227 10.6 0.521 2.445 8.224 9.182 mp_sum_l 807 5.4 5.431 8.358 5.431 8.358 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.207 7.888 mp_irecv_dv 3311 11.0 3.187 7.836 3.187 7.836 calculate_norms 2376 9.8 5.995 6.801 5.995 6.801 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.991 6.643 multiply_cannon_sync_h2d 1332 9.7 4.787 6.250 4.787 6.250 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.184 5.385 arnoldi_extremal 4 6.8 0.000 0.000 4.711 4.741 arnoldi_normal_ev 4 7.8 0.001 0.005 4.711 4.741 build_subspace 16 8.4 0.016 0.029 4.351 4.353 ls_scf_post 1 4.0 0.001 0.010 3.556 3.568 ls_scf_store_result 1 5.0 0.000 0.000 3.239 3.376 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.122 3.358 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.223 2.739 3.225 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.616 2.699 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.151 2.512 mp_allgather_i34 111 8.7 1.004 2.465 1.004 2.465 make_images_pack 222 9.7 2.042 2.402 2.044 2.404 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.187 2.328 rebuild_ks_matrix 3 7.3 0.000 0.000 2.174 2.315 qs_ks_build_kohn_sham_matrix 3 8.3 0.023 0.182 2.174 2.315 dbcsr_sort_data 436 11.2 1.841 2.087 1.841 2.087 dbcsr_data_new 4174 10.1 1.611 1.866 1.611 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.323000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 2.705023E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.051 94.966 94.967 qs_energies 1 2.0 0.000 0.000 94.402 94.405 ls_scf 1 3.0 0.000 0.000 92.902 92.905 dbcsr_multiply_generic 111 6.7 0.015 0.017 77.582 77.868 ls_scf_main 1 4.0 0.000 0.000 58.036 58.041 multiply_cannon 111 7.7 0.040 0.069 53.139 56.964 multiply_cannon_loop 111 8.7 0.099 0.104 49.603 54.036 density_matrix_trs4 2 5.0 0.002 0.003 52.060 52.226 mp_waitall_1 7281 11.0 24.199 34.395 24.199 34.395 ls_scf_init_scf 1 4.0 0.000 0.000 31.319 31.321 ls_scf_init_matrix_S 1 5.0 0.001 0.016 30.106 30.189 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.261 27.275 multiply_cannon_multrec 888 9.7 12.587 15.216 21.130 24.351 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.297 23.515 make_m2s 222 7.7 0.006 0.007 17.166 18.407 make_images 222 8.7 1.970 2.313 17.127 18.368 hybrid_alltoall_any 227 10.6 0.619 2.851 9.447 10.897 make_images_data 222 9.7 0.003 0.004 9.771 10.770 mp_sum_l 807 5.4 5.716 9.728 5.716 9.728 dbcsr_mm_accdrv_process 3754 10.4 0.239 0.412 8.082 9.256 dbcsr_mm_accdrv_process_sort 3754 11.4 7.668 8.844 7.668 8.844 multiply_cannon_sync_h2d 888 9.7 6.027 7.646 6.027 7.646 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.259 7.585 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.329 mp_irecv_dv 2335 11.1 2.457 7.286 2.457 7.286 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.753 6.755 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.240 5.448 arnoldi_extremal 4 6.8 0.000 0.000 5.034 5.053 arnoldi_normal_ev 4 7.8 0.001 0.004 5.034 5.053 calculate_norms 1584 9.8 4.311 4.760 4.311 4.760 build_subspace 16 8.4 0.014 0.020 4.734 4.740 mp_allgather_i34 111 8.7 1.375 3.825 1.375 3.825 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.418 3.750 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.588 3.017 3.590 ls_scf_post 1 4.0 0.000 0.000 3.547 3.553 ls_scf_store_result 1 5.0 0.000 0.000 3.289 3.376 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.865 2.985 make_images_sizes 222 9.7 0.000 0.000 1.051 2.248 mp_alltoall_i44 222 10.7 1.051 2.247 1.051 2.247 dbcsr_sort_data 325 11.1 1.903 2.180 1.903 2.180 make_images_pack 222 9.7 1.817 2.140 1.820 2.142 dbcsr_data_release 9322 10.9 1.308 1.948 1.308 1.948 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.932 1.934 rebuild_ks_matrix 3 7.3 0.000 0.000 1.914 1.916 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.914 1.916 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.967000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2194.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 3.297636E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.078 97.969 97.972 qs_energies 1 2.0 0.000 0.001 97.235 97.270 ls_scf 1 3.0 0.000 0.001 95.281 95.317 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.362 78.597 ls_scf_main 1 4.0 0.000 0.001 58.929 58.930 multiply_cannon 111 7.7 0.060 0.144 51.687 55.580 density_matrix_trs4 2 5.0 0.006 0.036 52.575 52.697 multiply_cannon_loop 111 8.7 0.114 0.126 46.681 49.771 ls_scf_init_scf 1 4.0 0.000 0.003 33.077 33.078 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.487 31.549 mp_waitall_1 6369 11.0 22.837 29.794 22.837 29.794 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.831 28.842 multiply_cannon_multrec 1332 9.7 14.169 17.244 22.053 25.187 make_m2s 222 7.7 0.006 0.007 21.195 22.560 make_images 222 8.7 3.138 3.609 21.145 22.512 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.335 16.572 make_images_data 222 9.7 0.004 0.004 11.863 13.479 hybrid_alltoall_any 227 10.6 0.798 3.829 11.201 13.000 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.410 7.525 8.992 dbcsr_mm_accdrv_process_sort 3641 11.4 7.103 8.578 7.103 8.578 mp_sum_l 807 5.4 3.894 7.105 3.894 7.105 multiply_cannon_sync_h2d 1332 9.7 5.533 6.259 5.533 6.259 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.083 6.014 mp_irecv_dv 3229 10.9 2.062 5.939 2.062 5.939 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.897 5.398 arnoldi_extremal 4 6.8 0.000 0.000 5.179 5.192 arnoldi_normal_ev 4 7.8 0.001 0.004 5.179 5.192 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.472 4.985 build_subspace 16 8.4 0.014 0.021 4.845 4.854 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.691 4.836 calculate_norms 2376 9.8 4.193 4.519 4.193 4.519 mp_allgather_i34 111 8.7 2.120 4.280 2.120 4.280 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.566 3.866 dbcsr_matrix_vector_mult_local 304 10.0 3.177 3.684 3.179 3.686 dbcsr_sort_data 658 11.4 3.108 3.550 3.108 3.550 ls_scf_post 1 4.0 0.000 0.000 3.275 3.312 ls_scf_dm_to_ks 2 5.0 0.030 0.249 3.265 3.307 dbcsr_special_finalize 555 9.7 0.006 0.007 2.847 3.237 dbcsr_merge_single_wm 555 10.7 0.536 0.668 2.838 3.230 ls_scf_store_result 1 5.0 0.000 0.000 2.949 3.027 dbcsr_data_release 10477 10.7 1.587 2.466 1.587 2.466 dbcsr_finalize 304 7.8 0.049 0.061 1.810 2.059 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.983 1.985 rebuild_ks_matrix 3 7.3 0.000 0.000 1.960 1.962 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.960 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.972000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2719.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 4.643500E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.057 99.962 99.963 qs_energies 1 2.0 0.000 0.000 98.807 98.813 ls_scf 1 3.0 0.000 0.000 96.850 96.860 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.853 78.045 ls_scf_main 1 4.0 0.000 0.000 62.210 62.211 multiply_cannon 111 7.7 0.079 0.184 55.357 60.847 density_matrix_trs4 2 5.0 0.002 0.003 54.937 55.036 multiply_cannon_loop 111 8.7 0.069 0.076 50.977 52.482 mp_waitall_1 5436 11.0 26.729 32.371 26.729 32.371 ls_scf_init_scf 1 4.0 0.000 0.000 31.036 31.041 ls_scf_init_matrix_S 1 5.0 0.002 0.034 29.323 29.365 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.112 27.124 multiply_cannon_multrec 444 9.7 14.031 16.700 21.060 22.706 make_m2s 222 7.7 0.004 0.005 17.812 20.181 make_images 222 8.7 3.719 4.417 17.750 20.121 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.103 16.218 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.373 14.952 make_images_data 222 9.7 0.003 0.004 10.050 12.406 hybrid_alltoall_any 227 10.6 0.788 3.756 9.905 12.264 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.340 6.732 7.835 dbcsr_mm_accdrv_process_sort 3003 11.4 6.413 7.491 6.413 7.491 multiply_cannon_sync_h2d 444 9.7 6.566 7.467 6.566 7.467 mp_allgather_i34 111 8.7 2.614 7.005 2.614 7.005 arnoldi_extremal 4 6.8 0.000 0.000 5.752 5.766 arnoldi_normal_ev 4 7.8 0.001 0.004 5.752 5.766 build_subspace 16 8.4 0.015 0.020 5.369 5.376 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.619 4.733 mp_sum_l 807 5.4 2.725 4.482 2.725 4.482 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.685 4.382 mp_irecv_dv 1241 11.2 1.666 4.361 1.666 4.361 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.142 4.351 dbcsr_matrix_vector_mult_local 304 10.0 3.679 4.155 3.681 4.157 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.678 3.786 calculate_norms 792 9.8 3.559 3.657 3.559 3.657 ls_scf_post 1 4.0 0.000 0.000 3.604 3.610 make_images_sizes 222 9.7 0.000 0.000 1.125 3.498 mp_alltoall_i44 222 10.7 1.125 3.497 1.125 3.497 ls_scf_store_result 1 5.0 0.000 0.000 3.387 3.425 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.818 3.278 dbcsr_finalize 304 7.8 0.062 0.078 2.195 2.249 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.225 2.226 rebuild_ks_matrix 3 7.3 0.000 0.000 2.191 2.192 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.191 2.192 dbcsr_merge_all 275 8.9 0.472 0.521 2.043 2.092 dbcsr_data_release 10123 10.8 1.330 2.020 1.330 2.020 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.963000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3687.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b3d50053968d9441899c157df378c42aede3ea5_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 8.717431E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.105 109.036 109.037 qs_energies 1 2.0 0.000 0.000 107.523 107.541 ls_scf 1 3.0 0.000 0.000 104.271 104.289 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.647 77.776 ls_scf_main 1 4.0 0.000 0.000 66.050 66.050 density_matrix_trs4 2 5.0 0.002 0.003 56.752 56.813 multiply_cannon 111 7.7 0.130 0.245 50.109 52.134 multiply_cannon_loop 111 8.7 0.068 0.070 46.623 47.595 ls_scf_init_scf 1 4.0 0.000 0.000 34.545 34.546 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.875 32.894 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.725 29.737 mp_waitall_1 4527 11.1 22.270 26.086 22.270 26.086 make_m2s 222 7.7 0.005 0.005 23.807 24.803 make_images 222 8.7 4.583 4.976 23.701 24.695 multiply_cannon_multrec 444 9.7 17.850 18.698 22.519 23.095 hybrid_alltoall_any 227 10.6 1.658 3.626 12.890 15.540 make_images_data 222 9.7 0.003 0.003 13.067 15.493 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.587 11.160 multiply_cannon_sync_h2d 444 9.7 8.851 8.902 8.851 8.902 arnoldi_extremal 4 6.8 0.000 0.000 7.332 7.345 arnoldi_normal_ev 4 7.8 0.002 0.009 7.332 7.345 build_subspace 16 8.4 0.026 0.036 6.787 6.797 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.440 5.580 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.265 5.524 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.262 5.351 dbcsr_matrix_vector_mult_local 304 10.0 5.017 5.317 5.020 5.320 dbcsr_mm_accdrv_process 1814 10.4 0.239 0.449 4.504 4.745 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.253 4.122 4.253 ls_scf_post 1 4.0 0.000 0.000 3.676 3.694 mp_allgather_i34 111 8.7 1.134 3.568 1.134 3.568 make_images_sizes 222 9.7 0.000 0.000 1.423 3.469 mp_alltoall_i44 222 10.7 1.423 3.468 1.423 3.468 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.424 calculate_norms 792 9.8 3.237 3.274 3.237 3.274 qs_energies_init_hamiltonians 1 3.0 0.315 0.321 3.220 3.220 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.161 dbcsr_merge_all 275 8.9 0.888 0.917 2.863 2.941 dbcsr_complete_redistribute 5 7.6 1.432 1.461 2.766 2.890 dbcsr_data_release 12724 10.6 2.330 2.857 2.330 2.857 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.591 2.592 matrix_ls_to_qs 2 6.0 0.000 0.000 2.418 2.555 rebuild_ks_matrix 3 7.3 0.000 0.000 2.525 2.526 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.525 2.526 dbcsr_sort_data 325 11.1 2.438 2.496 2.438 2.496 dbcsr_new_transposed 4 7.5 0.244 0.252 2.292 2.308 dbcsr_frobenius_norm 74 6.6 2.056 2.130 2.190 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.205 dbcsr_add_anytype 103 7.2 0.859 0.893 2.131 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.037000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6818.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6b3d50053968d9441899c157df378c42aede3ea5 Summary: empty Status: OK