=== 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: 186428f4e848f883d21da2d2b78d036968ebedff ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/01 job id: 47197344 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/02 job id: 47197345 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/03 job id: 47197346 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/04 job id: 47197347 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/05 job id: 47197348 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/06 job id: 47197349 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/07 job id: 47197350 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/08 job id: 47197351 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/09 job id: 47197352 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/10 job id: 47197353 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/11 job id: 47197354 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/12 job id: 47197355 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/13 job id: 47197356 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/14 job id: 47197358 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/15 job id: 47197359 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/16 job id: 47197360 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/17 job id: 47197362 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/18 job id: 47197363 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/19 job id: 47197364 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/20 job id: 47197366 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/21 job id: 47197367 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/22 job id: 47197368 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/23 job id: 47197369 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/24 job id: 47197370 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/25 job id: 47197371 --- 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/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/26 job id: 47197372 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/27 job id: 47197374 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.041 146.545 146.546 farming_run 1 2.0 145.927 145.929 146.488 146.491 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464390E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 120.072 120.073 qs_energies 1 2.0 0.000 0.000 119.753 119.757 mp2_main 1 3.0 0.000 0.000 115.819 115.822 mp2_gpw_main 1 4.0 0.019 0.025 113.857 113.861 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 93.025 93.442 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.972 55.389 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.415 46.629 get_2c_integrals 1 6.0 0.008 0.009 37.266 37.881 integrate_v_rspace 273 8.0 0.432 0.444 24.821 29.885 pw_transfer 6555 10.6 0.376 0.384 27.440 28.004 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.128 26.710 grid_integrate_task_list 273 9.0 20.599 26.169 20.599 26.169 fft_wrap_pw1pw2_100 2178 12.4 1.156 1.209 23.600 24.188 rpa_ri_compute_en 1 5.0 0.020 0.024 20.717 20.976 compute_2c_integrals 1 7.0 0.002 0.002 19.672 19.673 cp_fm_cholesky_decompose 12 8.2 18.272 19.402 18.272 19.402 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.802 18.908 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.437 18.800 18.907 cholesky_decomp 1 7.0 0.000 0.000 16.442 17.041 fft3d_s 5443 13.4 16.176 16.645 16.198 16.666 ao_to_mo_and_store_B_mult_1 272 7.0 10.724 15.305 10.724 15.305 calculate_wavefunction 272 8.0 5.409 5.548 12.547 13.151 rpa_num_int 1 6.0 0.000 0.000 11.849 11.849 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 11.286 11.847 calc_mat_Q 8 8.0 0.000 0.000 9.506 9.623 contract_S_to_Q 8 9.0 0.000 0.000 8.933 9.053 parallel_gemm_fm 14 9.1 0.000 0.000 8.531 8.645 parallel_gemm_fm_cosma 14 10.1 8.531 8.645 8.531 8.645 calc_potential_gpw 544 9.5 0.005 0.005 8.217 8.524 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.159 8.378 create_integ_mat 1 6.0 0.022 0.028 8.353 8.354 potential_pw2rs 545 10.0 0.107 0.109 7.704 8.325 collocate_single_gaussian 272 10.0 0.039 0.041 7.440 7.652 array2fm 1 7.0 0.000 0.000 6.693 7.313 pw_scatter_s 2720 13.7 4.450 4.599 4.450 4.599 pw_gather_s 2722 13.2 3.889 4.306 3.889 4.306 scf_env_do_scf 1 3.0 0.000 0.000 3.569 3.569 init_scf_loop 1 4.0 0.000 0.000 3.373 3.373 cp_fm_syevd 4 7.5 0.000 0.000 3.228 3.233 array2fm_buffer_send 1 8.0 2.904 3.137 2.904 3.137 prepare_preconditioner 1 5.0 0.000 0.000 3.056 3.057 make_preconditioner 1 6.0 0.000 0.000 2.980 2.981 cp_fm_redistribute_end 4 8.5 1.474 2.943 1.475 2.944 cp_fm_syevd_base 4 8.5 1.466 2.933 1.466 2.933 make_full_all 1 7.0 0.000 0.000 2.899 2.901 mp_sync 25 8.8 1.180 2.448 1.180 2.448 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.861029, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2801.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.042 408.396 408.398 farming_run 1 2.0 407.381 407.392 408.307 408.311 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224126E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 13520837. 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 0.025 0.046 211.254 211.255 qs_energies 1 2.0 0.000 0.001 210.966 210.987 scf_env_do_scf 1 3.0 0.000 0.000 106.984 106.984 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.402 105.411 rebuild_ks_matrix 4 6.0 0.000 0.000 105.400 105.409 qs_ks_build_kohn_sham_matrix 4 7.0 0.065 0.075 105.400 105.409 hfx_ks_matrix 4 8.0 0.001 0.001 104.844 104.849 integrate_four_center 4 9.0 0.143 0.447 104.843 104.848 mp2_main 1 3.0 0.001 0.016 103.635 103.655 mp2_gpw_main 1 4.0 0.032 0.047 101.876 101.897 integrate_four_center_main 4 10.0 0.093 0.481 96.724 99.231 integrate_four_center_bin 267 11.0 96.632 98.749 96.632 98.749 init_scf_loop 1 4.0 0.000 0.000 92.419 92.419 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 74.587 75.674 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.940 55.025 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.161 41.709 46.832 integrate_v_rspace 95 8.0 0.396 0.572 28.243 33.145 pw_transfer 2240 10.6 0.144 0.160 30.004 30.444 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 29.008 29.482 mp2_ri_gpw_compute_en 1 5.0 0.057 0.081 27.114 28.793 grid_integrate_task_list 95 9.0 23.379 28.446 23.379 28.446 ao_to_mo_and_store_B_mult_1 91 7.0 10.556 28.292 10.556 28.292 fft_wrap_pw1pw2_100 730 12.4 1.296 1.441 26.632 27.082 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.884 25.384 25.394 get_2c_integrals 1 6.0 0.000 0.000 20.394 20.582 compute_2c_integrals 1 7.0 0.002 0.003 19.342 19.348 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.912 19.171 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.864 18.911 19.170 fft3d_s 1823 13.4 18.417 18.732 18.430 18.746 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.560 14.560 calculate_wavefunction 91 8.0 2.038 2.069 9.750 9.965 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.594 8.852 9.356 potential_pw2rs 186 10.0 0.034 0.035 8.794 9.326 mp2_ri_gpw_compute_en_comm 22 7.0 0.511 0.532 8.306 8.905 local_gemm 172 8.0 8.296 8.777 8.296 8.777 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.269 8.643 collocate_single_gaussian 91 10.0 0.017 0.021 7.909 8.175 calc_potential_gpw 182 9.5 0.002 0.002 7.920 8.095 mp_sendrecv_dm3 2068 8.0 6.307 6.885 6.307 6.885 mp_sync 37 10.5 2.849 6.654 2.849 6.654 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.407 6.348 6.407 pw_gather_s 912 13.2 4.886 5.338 4.886 5.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.874908, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 451.796992E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 572054. 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 0.021 0.044 67.860 67.861 qs_mol_dyn_low 1 2.0 0.003 0.005 67.512 67.520 qs_forces 11 3.9 0.010 0.048 67.441 67.444 qs_energies 11 4.9 0.003 0.011 65.521 65.540 scf_env_do_scf 11 5.9 0.001 0.036 58.190 58.190 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 54.619 54.621 qs_scf_new_mos 108 7.5 0.000 0.001 39.261 39.589 qs_scf_loop_do_ot 108 8.5 0.000 0.001 39.260 39.589 dbcsr_multiply_generic 2286 12.5 0.095 0.099 37.415 37.883 ot_scf_mini 108 9.5 0.002 0.002 37.459 37.642 velocity_verlet 10 3.0 0.001 0.001 33.225 33.226 multiply_cannon 2286 13.5 0.187 0.200 27.313 28.754 multiply_cannon_loop 2286 14.5 1.490 1.585 25.812 27.340 ot_mini 108 10.5 0.001 0.001 22.070 22.327 qs_ot_get_derivative 108 11.5 0.001 0.001 18.903 19.084 mp_waitall_1 245248 16.5 10.552 16.426 10.552 16.426 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.139 13.129 multiply_cannon_multrec 54864 15.5 4.141 6.371 7.751 11.219 rebuild_ks_matrix 119 8.3 0.000 0.000 10.804 10.950 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 10.803 10.949 qs_ot_get_p 119 10.4 0.001 0.001 10.155 10.431 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.537 9.670 mp_sum_l 7287 12.8 5.823 7.586 5.823 7.586 qs_rho_update_rho_low 119 7.7 0.001 0.001 7.380 7.515 calculate_rho_elec 119 8.7 0.012 0.017 7.380 7.514 multiply_cannon_sync_h2d 54864 15.5 5.718 7.047 5.718 7.047 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.566 7.018 sum_up_and_integrate 119 10.3 0.012 0.014 6.750 6.761 integrate_v_rspace 119 11.3 0.002 0.003 6.738 6.750 qs_ot_p2m_diag 50 11.0 0.004 0.007 6.405 6.445 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.307 6.409 init_scf_run 11 5.9 0.000 0.001 5.756 5.757 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.756 5.756 rs_pw_transfer 974 11.9 0.011 0.012 5.305 5.508 cp_dbcsr_syevd 50 12.0 0.003 0.003 5.482 5.484 density_rs2pw 119 9.7 0.004 0.005 4.939 5.082 dbcsr_mm_accdrv_process 76910 16.1 1.191 1.894 3.532 4.943 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.852 4.857 cp_fm_redistribute_end 50 14.0 2.481 4.772 2.495 4.782 cp_fm_diag_elpa_base 50 14.0 2.269 4.561 2.275 4.572 pw_transfer 1439 11.6 0.053 0.058 4.373 4.543 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 4.296 4.470 make_m2s 4572 13.5 0.055 0.056 3.976 4.169 fft3d_ps 1201 14.6 0.378 0.499 3.947 4.127 make_images 4572 14.5 0.134 0.140 3.890 4.079 potential_pw2rs 119 12.3 0.004 0.004 3.839 3.926 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 2.128 3.854 fft_wrap_pw1pw2_140 487 13.2 0.176 0.196 3.308 3.481 init_scf_loop 11 6.9 0.000 0.000 3.472 3.474 calculate_dm_sparse 119 9.5 0.000 0.001 3.228 3.395 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.064 3.262 apply_single 119 13.6 0.000 0.000 3.063 3.262 jit_kernel_multiply 13 15.8 2.279 2.960 2.279 2.960 mp_alltoall_d11v 2130 13.8 2.703 2.893 2.703 2.893 ot_diis_step 108 11.5 0.006 0.006 2.887 2.888 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.819 2.876 calculate_first_density_matrix 1 7.0 0.000 0.000 2.853 2.859 mp_alltoall_z22v 1201 16.6 2.660 2.792 2.660 2.792 wfi_extrapolate 11 7.9 0.001 0.001 2.762 2.762 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.752 2.755 make_images_sizes 4572 15.5 0.004 0.004 2.179 2.656 mp_alltoall_i44 4572 16.5 2.175 2.652 2.175 2.652 acc_transpose_blocks 54864 15.5 0.226 0.248 1.843 2.338 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.196 2.249 mp_sum_d 4135 12.0 1.530 2.131 1.530 2.131 grid_integrate_task_list 119 12.3 1.992 2.120 1.992 2.120 prepare_preconditioner 11 7.9 0.000 0.000 2.032 2.061 make_preconditioner 11 8.9 0.000 0.000 2.032 2.060 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.915 1.955 make_images_data 4572 15.5 0.039 0.043 1.103 1.770 mp_waitany 12084 13.8 1.644 1.766 1.644 1.766 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.670 1.694 yz_to_x 249 15.8 0.020 0.029 1.569 1.694 mp_allgather_i34 2286 14.5 1.049 1.643 1.049 1.643 mp_sendrecv_dv 22610 12.7 1.341 1.443 1.341 1.443 grid_collocate_task_list 119 9.7 1.357 1.439 1.357 1.439 hybrid_alltoall_any 4725 16.4 0.041 0.114 0.934 1.418 arnoldi_extremal 119 11.4 0.001 0.001 1.260 1.380 arnoldi_normal_ev 119 12.4 0.002 0.002 1.259 1.378 dbcsr_complete_redistribute 329 12.2 0.049 0.082 1.332 1.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=67.861000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 489.406464E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1492288. 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 0.013 0.032 41.275 41.276 qs_mol_dyn_low 1 2.0 0.003 0.003 41.073 41.081 qs_forces 11 3.9 0.003 0.003 41.006 41.007 qs_energies 11 4.9 0.001 0.002 39.245 39.251 scf_env_do_scf 11 5.9 0.001 0.002 33.785 33.786 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 31.065 31.066 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.117 22.514 qs_scf_new_mos 108 7.5 0.001 0.001 20.979 21.214 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.979 21.214 ot_scf_mini 108 9.5 0.002 0.003 20.046 20.213 velocity_verlet 10 3.0 0.001 0.002 19.655 19.656 multiply_cannon 2286 13.5 0.208 0.215 16.658 18.296 multiply_cannon_loop 2286 14.5 0.910 0.980 15.423 16.922 ot_mini 108 10.5 0.001 0.001 12.235 12.469 mp_waitall_1 200699 16.5 6.132 11.326 6.132 11.326 qs_ot_get_derivative 108 11.5 0.001 0.001 9.712 9.886 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.202 9.433 multiply_cannon_multrec 27432 15.5 1.990 4.395 6.008 8.704 rebuild_ks_matrix 119 8.3 0.000 0.000 7.711 7.844 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.711 7.843 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.817 6.937 dbcsr_mm_accdrv_process 47894 16.0 3.063 5.272 3.948 5.810 qs_ot_get_p 119 10.4 0.001 0.001 4.797 5.022 sum_up_and_integrate 119 10.3 0.025 0.029 4.652 4.658 integrate_v_rspace 119 11.3 0.002 0.003 4.627 4.635 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.744 4.596 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.341 4.381 calculate_rho_elec 119 8.7 0.021 0.024 4.340 4.381 init_scf_run 11 5.9 0.000 0.001 4.208 4.209 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.208 4.208 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.093 4.167 apply_single 119 13.6 0.000 0.000 3.093 4.166 mp_sum_l 7287 12.8 2.194 4.155 2.194 4.155 rs_pw_transfer 974 11.9 0.010 0.013 3.228 3.689 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.185 3.203 make_m2s 4572 13.5 0.052 0.054 2.820 3.096 density_rs2pw 119 9.7 0.004 0.004 2.591 3.077 make_images 4572 14.5 0.201 0.237 2.730 3.004 multiply_cannon_sync_h2d 27432 15.5 2.161 2.862 2.161 2.862 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.722 2.723 init_scf_loop 11 6.9 0.000 0.000 2.688 2.688 calculate_first_density_matrix 1 7.0 0.000 0.000 2.527 2.529 ot_diis_step 108 11.5 0.011 0.011 2.468 2.469 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.275 2.365 potential_pw2rs 119 12.3 0.006 0.006 2.329 2.352 calculate_dm_sparse 119 9.5 0.000 0.001 2.223 2.300 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.287 2.290 cp_fm_redistribute_end 50 14.0 1.159 2.251 1.163 2.255 pw_transfer 1439 11.6 0.066 0.071 2.195 2.250 cp_fm_diag_elpa_base 50 14.0 1.058 2.145 1.087 2.187 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.103 2.160 jit_kernel_multiply 10 16.0 0.833 2.067 0.833 2.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.980 1.982 grid_integrate_task_list 119 12.3 1.832 1.927 1.832 1.927 make_images_data 4572 15.5 0.045 0.051 1.256 1.777 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.723 1.762 prepare_preconditioner 11 7.9 0.000 0.000 1.725 1.753 make_preconditioner 11 8.9 0.000 0.000 1.725 1.753 fft3d_ps 1201 14.6 0.524 0.580 1.654 1.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.617 1.678 wfi_extrapolate 11 7.9 0.001 0.001 1.615 1.615 fft_wrap_pw1pw2_140 487 13.2 0.203 0.213 1.554 1.606 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.109 1.584 mp_alltoall_d11v 2130 13.8 1.388 1.571 1.388 1.571 acc_transpose_blocks 27432 15.5 0.109 0.114 1.214 1.520 mp_allgather_i34 2286 14.5 0.675 1.499 0.675 1.499 grid_collocate_task_list 119 9.7 1.272 1.422 1.272 1.422 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.412 1.421 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.265 1.317 mp_sum_d 4135 12.0 0.620 1.104 0.620 1.104 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.564 1.032 mp_waitany 5720 13.7 0.546 1.032 0.546 1.032 make_images_sizes 4572 15.5 0.005 0.005 0.734 1.019 mp_alltoall_i44 4572 16.5 0.729 1.015 0.729 1.015 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.955 0.968 rs_pw_transfer_PW2RS_50 119 14.3 0.593 0.612 0.858 0.950 mp_alltoall_z22v 1201 16.6 0.857 0.947 0.857 0.947 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.945 0.946 acc_transpose_blocks_kernels 27432 16.5 0.186 0.278 0.665 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.276000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.000000, yerr=1.537412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.547776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1049484. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.050 43.631 43.637 qs_mol_dyn_low 1 2.0 0.003 0.003 43.023 43.033 qs_forces 11 3.9 0.002 0.002 42.955 42.957 qs_energies 11 4.9 0.003 0.015 41.108 41.113 scf_env_do_scf 11 5.9 0.001 0.003 35.562 35.564 scf_env_do_scf_inner_loop 108 6.5 0.003 0.011 31.771 31.773 dbcsr_multiply_generic 2286 12.5 0.099 0.103 20.347 20.953 velocity_verlet 10 3.0 0.001 0.001 20.628 20.629 qs_scf_new_mos 108 7.5 0.001 0.001 20.575 20.611 qs_scf_loop_do_ot 108 8.5 0.001 0.002 20.574 20.611 ot_scf_mini 108 9.5 0.002 0.004 19.613 19.637 multiply_cannon 2286 13.5 0.199 0.205 14.577 16.020 multiply_cannon_loop 2286 14.5 0.643 0.675 13.246 15.046 ot_mini 108 10.5 0.001 0.001 11.598 11.644 qs_ot_get_derivative 108 11.5 0.001 0.001 9.666 9.693 rebuild_ks_matrix 119 8.3 0.000 0.000 8.172 8.318 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.171 8.317 mp_waitall_1 158411 16.6 5.106 7.926 5.106 7.926 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.228 7.361 multiply_cannon_multrec 18288 15.5 1.914 2.859 6.951 7.260 dbcsr_mm_accdrv_process 38222 16.0 4.249 5.834 4.954 5.891 sum_up_and_integrate 119 10.3 0.031 0.032 5.524 5.531 integrate_v_rspace 119 11.3 0.002 0.003 5.493 5.503 qs_ot_get_p 119 10.4 0.001 0.001 5.344 5.382 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.103 5.148 calculate_rho_elec 119 8.7 0.031 0.031 5.102 5.147 rs_pw_transfer 974 11.9 0.009 0.010 4.575 4.798 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 2.097 4.709 init_scf_run 11 5.9 0.000 0.001 4.226 4.226 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.225 4.226 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.483 4.119 init_scf_loop 11 6.9 0.001 0.004 3.753 3.755 density_rs2pw 119 9.7 0.004 0.004 3.474 3.721 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.480 3.498 make_m2s 4572 13.5 0.045 0.045 3.057 3.496 make_images 4572 14.5 0.193 0.205 2.971 3.409 potential_pw2rs 119 12.3 0.007 0.009 3.166 3.203 cp_dbcsr_syevd 50 12.0 0.003 0.004 3.073 3.074 mp_sum_l 7287 12.8 2.375 2.918 2.375 2.918 prepare_preconditioner 11 7.9 0.000 0.001 2.761 2.765 make_preconditioner 11 8.9 0.000 0.002 2.761 2.765 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.256 2.710 apply_single 119 13.6 0.000 0.000 2.256 2.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.581 2.686 pw_transfer 1439 11.6 0.066 0.070 2.617 2.665 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.610 2.620 cp_fm_diag_elpa_base 50 14.0 2.577 2.598 2.606 2.617 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.524 2.573 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.524 2.535 calculate_first_density_matrix 1 7.0 0.000 0.003 2.476 2.477 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.158 fft3d_ps 1201 14.6 0.535 0.555 2.014 2.070 jit_kernel_multiply 10 16.0 0.653 1.997 0.653 1.997 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.988 1.990 grid_integrate_task_list 119 12.3 1.791 1.916 1.791 1.916 make_images_data 4572 15.5 0.045 0.048 1.377 1.884 ot_diis_step 108 11.5 0.011 0.011 1.835 1.836 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.686 1.727 fft_wrap_pw1pw2_140 487 13.2 0.250 0.258 1.677 1.716 wfi_extrapolate 11 7.9 0.001 0.001 1.669 1.669 hybrid_alltoall_any 4725 16.4 0.055 0.114 1.161 1.656 multiply_cannon_sync_h2d 18288 15.5 1.248 1.480 1.248 1.480 mp_alltoall_d11v 2130 13.8 1.209 1.437 1.209 1.437 grid_collocate_task_list 119 9.7 1.247 1.399 1.247 1.399 mp_sendrecv_dv 11067 12.7 1.325 1.369 1.325 1.369 mp_allgather_i34 2286 14.5 0.783 1.356 0.783 1.356 make_images_sizes 4572 15.5 0.005 0.005 0.977 1.355 acc_transpose_blocks 18288 15.5 0.076 0.079 1.314 1.350 mp_alltoall_i44 4572 16.5 0.972 1.350 0.972 1.350 mp_alltoall_z22v 1201 16.6 1.232 1.339 1.232 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.301 1.311 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.231 1.276 cp_fm_cholesky_invert 11 10.9 1.241 1.249 1.241 1.249 rs_pw_transfer_PW2RS_50 119 14.3 0.400 0.412 1.001 1.118 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.027 1.041 rs_pw_transfer_RS2PW_50 119 11.7 0.241 0.250 0.902 0.997 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.989 0.991 dbcsr_complete_redistribute 329 12.2 0.092 0.104 0.803 0.929 mp_waitany 9880 13.7 0.681 0.926 0.681 0.926 mp_sum_d 4135 12.0 0.717 0.909 0.717 0.909 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.567 0.907 acc_transpose_blocks_kernels 18288 16.5 0.218 0.228 0.858 0.882 cp_fm_cholesky_decompose 22 10.9 0.871 0.875 0.871 0.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=43.637000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=2.452861 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 550.502400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 47.771 47.772 qs_mol_dyn_low 1 2.0 0.003 0.003 47.451 47.460 qs_forces 11 3.9 0.006 0.032 47.381 47.382 qs_energies 11 4.9 0.002 0.007 45.349 45.356 scf_env_do_scf 11 5.9 0.001 0.001 39.348 39.349 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 33.900 33.901 velocity_verlet 10 3.0 0.001 0.002 24.293 24.295 qs_scf_new_mos 108 7.5 0.001 0.001 22.214 22.286 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.214 22.285 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.822 21.976 ot_scf_mini 108 9.5 0.003 0.004 21.077 21.138 multiply_cannon 2286 13.5 0.226 0.267 15.687 16.439 multiply_cannon_loop 2286 14.5 0.954 0.983 14.183 14.675 ot_mini 108 10.5 0.001 0.001 11.944 12.025 qs_ot_get_derivative 108 11.5 0.001 0.001 9.810 9.870 multiply_cannon_multrec 27432 15.5 2.349 3.028 8.632 8.942 rebuild_ks_matrix 119 8.3 0.000 0.000 8.555 8.628 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.554 8.627 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.640 7.709 dbcsr_mm_accdrv_process 47916 15.9 5.440 7.090 6.192 7.380 qs_ot_get_p 119 10.4 0.001 0.001 5.974 6.077 sum_up_and_integrate 119 10.3 0.036 0.039 5.412 5.418 init_scf_loop 11 6.9 0.001 0.006 5.399 5.401 integrate_v_rspace 119 11.3 0.003 0.003 5.376 5.383 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.265 5.296 calculate_rho_elec 119 8.7 0.040 0.046 5.264 5.295 mp_waitall_1 137007 16.6 4.464 5.279 4.464 5.279 rs_pw_transfer 974 11.9 0.009 0.010 4.355 4.525 init_scf_run 11 5.9 0.000 0.001 4.409 4.409 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.408 4.409 prepare_preconditioner 11 7.9 0.000 0.001 4.300 4.309 make_preconditioner 11 8.9 0.000 0.002 4.300 4.309 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.811 4.198 make_m2s 4572 13.5 0.054 0.056 3.834 4.112 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.490 4.081 make_images 4572 14.5 0.271 0.333 3.725 3.999 qs_ot_p2m_diag 50 11.0 0.016 0.023 3.895 3.907 density_rs2pw 119 9.7 0.004 0.004 3.539 3.701 cp_dbcsr_syevd 50 12.0 0.003 0.004 3.451 3.452 potential_pw2rs 119 12.3 0.009 0.010 3.023 3.035 pw_transfer 1439 11.6 0.066 0.070 2.893 2.935 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.461 2.933 apply_single 119 13.6 0.000 0.000 2.461 2.933 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.901 2.910 cp_fm_diag_elpa_base 50 14.0 2.843 2.865 2.896 2.905 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.801 2.846 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.764 2.798 mp_sum_l 7287 12.8 1.900 2.427 1.900 2.427 calculate_first_density_matrix 1 7.0 0.000 0.002 2.346 2.348 calculate_dm_sparse 119 9.5 0.000 0.000 2.263 2.331 fft3d_ps 1201 14.6 0.561 0.609 2.241 2.278 make_images_data 4572 15.5 0.045 0.048 1.757 2.213 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.094 2.098 ot_diis_step 108 11.5 0.012 0.012 2.071 2.071 fft_wrap_pw1pw2_140 487 13.2 0.287 0.301 2.015 2.061 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 1.260 2.019 wfi_extrapolate 11 7.9 0.001 0.001 1.981 1.981 hybrid_alltoall_any 4725 16.4 0.062 0.150 1.483 1.952 grid_integrate_task_list 119 12.3 1.808 1.908 1.808 1.908 jit_kernel_multiply 10 15.9 0.689 1.841 0.689 1.841 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.697 1.731 cp_fm_cholesky_invert 11 10.9 1.714 1.720 1.714 1.720 dbcsr_complete_redistribute 329 12.2 0.160 0.199 1.371 1.682 mp_alltoall_z22v 1201 16.6 1.563 1.597 1.563 1.597 mp_alltoall_d11v 2130 13.8 1.436 1.581 1.436 1.581 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.488 1.516 acc_transpose_blocks 27432 15.5 0.111 0.113 1.483 1.505 make_images_sizes 4572 15.5 0.005 0.005 1.056 1.481 mp_alltoall_i44 4572 16.5 1.051 1.476 1.051 1.476 grid_collocate_task_list 119 9.7 1.250 1.372 1.250 1.372 cp_fm_upper_to_full 72 14.2 0.929 1.356 0.929 1.356 mp_sendrecv_dv 8211 12.7 1.333 1.352 1.333 1.352 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.296 1.308 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.904 1.210 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.168 1.169 cp_fm_cholesky_decompose 22 10.9 1.042 1.048 1.042 1.048 mp_allgather_i34 2286 14.5 0.793 1.029 0.793 1.029 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.015 1.023 multiply_cannon_sync_h2d 27432 15.5 0.926 0.994 0.926 0.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=47.772000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=521.818182, yerr=4.281422 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 600.850432E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 37.130 37.130 qs_mol_dyn_low 1 2.0 0.003 0.003 36.888 36.896 qs_forces 11 3.9 0.002 0.003 36.822 36.823 qs_energies 11 4.9 0.001 0.002 34.911 34.915 scf_env_do_scf 11 5.9 0.001 0.002 29.324 29.324 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.627 25.628 velocity_verlet 10 3.0 0.001 0.001 19.084 19.087 qs_scf_new_mos 108 7.5 0.001 0.001 15.166 15.212 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.165 15.211 dbcsr_multiply_generic 2286 12.5 0.092 0.098 15.043 15.177 ot_scf_mini 108 9.5 0.002 0.002 14.331 14.376 multiply_cannon 2286 13.5 0.233 0.242 10.861 11.594 multiply_cannon_loop 2286 14.5 0.334 0.346 9.720 9.978 ot_mini 108 10.5 0.001 0.001 7.526 7.574 rebuild_ks_matrix 119 8.3 0.000 0.000 7.483 7.537 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.483 7.537 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.679 6.727 multiply_cannon_multrec 9144 15.5 1.635 1.848 6.008 6.208 qs_ot_get_derivative 108 11.5 0.001 0.001 6.025 6.067 sum_up_and_integrate 119 10.3 0.038 0.041 4.899 4.906 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.885 4.905 calculate_rho_elec 119 8.7 0.060 0.061 4.885 4.904 integrate_v_rspace 119 11.3 0.003 0.003 4.861 4.869 qs_ot_get_p 119 10.4 0.001 0.001 4.622 4.671 dbcsr_mm_accdrv_process 12550 15.8 3.366 4.283 4.272 4.344 mp_waitall_1 115863 16.7 3.595 4.166 3.595 4.166 init_scf_run 11 5.9 0.000 0.001 3.999 3.999 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.998 3.998 init_scf_loop 11 6.9 0.000 0.000 3.657 3.658 rs_pw_transfer 974 11.9 0.008 0.008 3.241 3.387 density_rs2pw 119 9.7 0.004 0.004 3.031 3.183 make_m2s 4572 13.5 0.035 0.036 2.941 3.141 qs_ot_p2m_diag 50 11.0 0.022 0.023 3.076 3.078 make_images 4572 14.5 0.268 0.302 2.850 3.049 pw_transfer 1439 11.6 0.066 0.070 2.806 2.818 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.776 2.777 prepare_preconditioner 11 7.9 0.000 0.000 2.734 2.740 make_preconditioner 11 8.9 0.000 0.000 2.734 2.740 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.712 2.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.583 2.635 potential_pw2rs 119 12.3 0.010 0.011 2.424 2.432 calculate_first_density_matrix 1 7.0 0.000 0.000 2.362 2.363 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.270 2.278 cp_fm_diag_elpa_base 50 14.0 2.232 2.257 2.267 2.275 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.145 2.163 fft3d_ps 1201 14.6 0.568 0.580 2.045 2.063 grid_integrate_task_list 119 12.3 1.911 2.048 1.911 2.048 fft_wrap_pw1pw2_140 487 13.2 0.364 0.375 2.000 2.014 calculate_dm_sparse 119 9.5 0.000 0.000 1.943 1.975 make_images_data 4572 15.5 0.039 0.042 1.449 1.849 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.819 1.822 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.757 1.781 jit_kernel_multiply 9 15.5 0.868 1.763 0.868 1.763 hybrid_alltoall_any 4725 16.4 0.063 0.176 1.315 1.744 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.571 1.586 wfi_extrapolate 11 7.9 0.001 0.001 1.550 1.551 grid_collocate_task_list 119 9.7 1.357 1.485 1.357 1.485 ot_diis_step 108 11.5 0.012 0.013 1.465 1.467 cp_fm_cholesky_invert 11 10.9 1.443 1.447 1.443 1.447 mp_alltoall_d11v 2130 13.8 1.267 1.399 1.267 1.399 mp_alltoall_z22v 1201 16.6 1.350 1.383 1.350 1.383 make_images_sizes 4572 15.5 0.005 0.005 0.827 1.361 mp_alltoall_i44 4572 16.5 0.823 1.357 0.823 1.357 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.301 1.346 apply_single 119 13.6 0.000 0.000 1.301 1.346 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 1.129 1.279 qs_energies_init_hamiltonians 11 5.9 0.010 0.013 1.256 1.257 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.246 mp_allgather_i34 2286 14.5 0.466 1.087 0.466 1.087 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.967 0.981 mp_sum_l 7287 12.8 0.818 0.979 0.818 0.979 acc_transpose_blocks 9144 15.5 0.038 0.039 0.961 0.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.933 0.939 mp_sendrecv_dv 5355 12.7 0.909 0.922 0.909 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.865 0.918 dbcsr_complete_redistribute 329 12.2 0.160 0.170 0.846 0.896 cp_fm_cholesky_decompose 22 10.9 0.827 0.830 0.827 0.830 yz_to_x 606 15.1 0.053 0.055 0.780 0.801 arnoldi_extremal 119 11.4 0.002 0.002 0.766 0.787 arnoldi_normal_ev 119 12.4 0.002 0.002 0.764 0.785 multiply_cannon_sync_h2d 9144 15.5 0.694 0.753 0.694 0.753 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.702 0.752 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.726 0.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=37.130000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.909091, yerr=5.743843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 741.838848E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 43.270 43.271 qs_mol_dyn_low 1 2.0 0.003 0.003 43.036 43.043 qs_forces 11 3.9 0.002 0.002 42.972 42.973 qs_energies 11 4.9 0.002 0.002 40.931 40.933 scf_env_do_scf 11 5.9 0.001 0.001 35.051 35.051 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.065 27.066 velocity_verlet 10 3.0 0.001 0.001 24.247 24.253 dbcsr_multiply_generic 2286 12.5 0.099 0.102 18.169 18.384 qs_scf_new_mos 108 7.5 0.001 0.001 16.809 16.907 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.808 16.906 ot_scf_mini 108 9.5 0.002 0.002 15.693 15.788 multiply_cannon 2286 13.5 0.303 0.317 13.885 14.889 multiply_cannon_loop 2286 14.5 0.344 0.352 12.519 13.437 ot_mini 108 10.5 0.001 0.001 9.230 9.342 multiply_cannon_multrec 9144 15.5 3.444 4.799 8.739 8.878 init_scf_loop 11 6.9 0.000 0.000 7.953 7.956 rebuild_ks_matrix 119 8.3 0.000 0.000 7.592 7.726 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.591 7.726 qs_ot_get_derivative 108 11.5 0.001 0.001 7.160 7.257 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.863 6.985 prepare_preconditioner 11 7.9 0.000 0.000 6.956 6.969 make_preconditioner 11 8.9 0.000 0.000 6.956 6.969 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.551 6.842 dbcsr_mm_accdrv_process 12550 15.8 4.411 6.247 5.172 6.540 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.516 4.523 calculate_rho_elec 119 8.7 0.118 0.121 4.515 4.522 cp_fm_upper_to_full 72 14.2 3.172 4.518 3.172 4.518 sum_up_and_integrate 119 10.3 0.064 0.066 4.215 4.222 integrate_v_rspace 119 11.3 0.003 0.004 4.151 4.157 init_scf_run 11 5.9 0.000 0.001 3.835 3.835 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.835 3.835 qs_ot_get_p 119 10.4 0.001 0.001 3.630 3.765 mp_waitall_1 94719 16.7 2.654 3.662 2.654 3.662 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.592 3.027 pw_transfer 1439 11.6 0.068 0.069 2.863 2.870 make_m2s 4572 13.5 0.038 0.038 2.618 2.867 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.764 2.771 dbcsr_complete_redistribute 329 12.2 0.285 0.287 1.989 2.763 make_images 4572 14.5 0.351 0.382 2.495 2.744 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.244 2.507 apply_single 119 13.6 0.000 0.000 2.244 2.507 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.655 2.433 density_rs2pw 119 9.7 0.004 0.004 2.375 2.388 calculate_dm_sparse 119 9.5 0.000 0.000 2.346 2.363 fft_wrap_pw1pw2_140 487 13.2 0.618 0.624 2.283 2.292 calculate_first_density_matrix 1 7.0 0.000 0.000 2.270 2.271 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.236 2.238 mp_alltoall_i22 627 13.8 1.391 2.203 1.391 2.203 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.264 2.178 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.400 2.172 grid_integrate_task_list 119 12.3 2.074 2.105 2.074 2.105 ot_diis_step 108 11.5 0.014 0.014 2.048 2.049 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.945 1.946 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.826 1.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.843 1.844 fft3d_ps 1201 14.6 0.598 0.608 1.780 1.786 mp_sum_l 7287 12.8 1.132 1.780 1.132 1.780 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.760 1.761 rs_pw_transfer 974 11.9 0.009 0.009 1.671 1.699 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.294 1.635 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.608 1.608 cp_fm_diag_elpa_base 50 14.0 1.455 1.512 1.605 1.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.566 1.603 cp_fm_cholesky_invert 11 10.9 1.587 1.590 1.587 1.590 make_images_data 4572 15.5 0.042 0.045 1.279 1.587 potential_pw2rs 119 12.3 0.014 0.015 1.558 1.562 grid_collocate_task_list 119 9.7 1.519 1.535 1.519 1.535 wfi_extrapolate 11 7.9 0.001 0.001 1.493 1.493 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.385 1.392 mp_alltoall_d11v 2130 13.8 1.205 1.218 1.205 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.170 1.187 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.099 1.117 mp_alltoall_z22v 1201 16.6 1.046 1.061 1.046 1.061 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.059 multiply_cannon_sync_h2d 9144 15.5 1.042 1.046 1.042 1.046 jit_kernel_multiply 6 15.5 0.733 1.025 0.733 1.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.944 0.956 qs_create_task_list 11 7.9 0.000 0.000 0.935 0.947 generate_qs_task_list 11 8.9 0.367 0.386 0.935 0.947 acc_transpose_blocks 9144 15.5 0.038 0.039 0.932 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.271000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.818182, yerr=12.626222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.272000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1460355. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 98.138 98.139 qs_mol_dyn_low 1 2.0 0.003 0.003 97.854 97.864 qs_forces 11 3.9 0.003 0.003 97.780 97.782 qs_energies 11 4.9 0.005 0.006 94.575 94.596 scf_env_do_scf 11 5.9 0.001 0.001 84.417 84.420 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 76.470 76.471 qs_scf_new_mos 99 7.5 0.000 0.001 54.965 55.155 qs_scf_loop_do_ot 99 8.5 0.001 0.001 54.965 55.154 dbcsr_multiply_generic 2055 12.4 0.107 0.113 54.268 54.575 ot_scf_mini 99 9.5 0.002 0.002 52.395 52.497 velocity_verlet 10 3.0 0.001 0.002 50.468 50.469 multiply_cannon 2055 13.4 0.181 0.185 43.262 44.332 multiply_cannon_loop 2055 14.4 1.533 1.562 41.740 42.898 ot_mini 99 10.5 0.001 0.001 28.822 28.942 qs_ot_get_derivative 99 11.5 0.001 0.001 21.956 22.097 multiply_cannon_multrec 49320 15.4 12.168 12.836 17.125 17.836 rebuild_ks_matrix 110 8.3 0.000 0.000 17.014 17.207 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 17.014 17.207 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.949 15.118 qs_ot_get_p 110 10.4 0.001 0.001 14.520 14.648 mp_waitall_1 220248 16.4 13.327 14.400 13.327 14.400 multiply_cannon_sync_h2d 49320 15.4 9.964 10.395 9.964 10.395 qs_ot_p2m_diag 48 11.0 0.013 0.019 10.325 10.393 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.084 9.216 calculate_rho_elec 110 8.6 0.022 0.026 9.083 9.216 cp_dbcsr_syevd 48 12.0 0.003 0.003 9.163 9.164 sum_up_and_integrate 110 10.3 0.036 0.043 9.044 9.056 integrate_v_rspace 110 11.3 0.003 0.004 9.007 9.029 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.110 8.663 cp_fm_diag_elpa 48 13.0 0.000 0.000 8.144 8.166 cp_fm_diag_elpa_base 48 14.0 8.115 8.143 8.141 8.164 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.670 8.088 init_scf_loop 11 6.9 0.000 0.000 7.898 7.899 init_scf_run 11 5.9 0.000 0.001 7.844 7.844 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.844 7.844 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.281 7.754 apply_single 110 13.6 0.000 0.001 7.281 7.754 rs_pw_transfer 902 11.9 0.011 0.013 5.950 6.655 ot_diis_step 99 11.5 0.005 0.006 6.648 6.649 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.435 6.482 density_rs2pw 110 9.6 0.004 0.005 5.748 6.311 make_m2s 4110 13.4 0.064 0.067 5.741 5.884 make_images 4110 14.4 0.179 0.194 5.638 5.782 mp_sum_l 6594 12.7 4.811 5.777 4.811 5.777 pw_transfer 1331 11.6 0.057 0.066 5.462 5.635 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.372 5.543 prepare_preconditioner 11 7.9 0.000 0.000 5.448 5.465 make_preconditioner 11 8.9 0.000 0.000 5.448 5.465 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.183 5.245 dbcsr_mm_accdrv_process 87628 16.1 2.001 2.071 4.833 5.131 wfi_extrapolate 11 7.9 0.001 0.001 5.034 5.035 fft3d_ps 1111 14.6 0.786 0.889 4.569 4.722 potential_pw2rs 110 12.3 0.006 0.007 4.437 4.527 fft_wrap_pw1pw2_140 451 13.1 0.446 0.497 4.194 4.371 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 3.190 4.203 calculate_dm_sparse 110 9.5 0.001 0.001 3.962 4.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.796 3.803 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.672 3.733 grid_integrate_task_list 110 12.3 3.216 3.489 3.216 3.489 mp_alltoall_z22v 1111 16.6 3.212 3.436 3.212 3.436 mp_alltoall_d11v 2046 13.8 2.681 3.318 2.681 3.318 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.272 3.315 cp_fm_cholesky_invert 11 10.9 3.006 3.012 3.006 3.012 calculate_first_density_matrix 1 7.0 0.000 0.000 2.669 2.674 make_images_data 4110 15.4 0.044 0.047 2.376 2.650 mp_waitany 14300 13.8 2.036 2.623 2.036 2.623 jit_kernel_multiply 13 15.9 2.552 2.587 2.552 2.587 make_images_sizes 4110 15.4 0.004 0.004 1.834 2.414 mp_alltoall_i44 4110 16.4 1.830 2.410 1.830 2.410 hybrid_alltoall_any 4261 16.3 0.082 0.476 2.010 2.362 grid_collocate_task_list 110 9.6 2.154 2.356 2.154 2.356 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.201 2.225 acc_transpose_blocks 49320 15.4 0.211 0.219 2.122 2.169 mp_sum_d 3889 11.9 1.530 2.044 1.530 2.044 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=98.139000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=3.214122 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.976704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2505149. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.033 93.146 93.148 qs_mol_dyn_low 1 2.0 0.003 0.003 92.828 92.838 qs_forces 11 3.9 0.003 0.003 92.749 92.750 qs_energies 11 4.9 0.005 0.034 89.016 89.038 scf_env_do_scf 11 5.9 0.001 0.001 79.068 79.072 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.903 65.904 velocity_verlet 10 3.0 0.002 0.009 48.271 48.274 qs_scf_new_mos 99 7.5 0.001 0.001 45.465 45.610 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.464 45.610 dbcsr_multiply_generic 2055 12.4 0.119 0.123 44.591 45.373 ot_scf_mini 99 9.5 0.003 0.003 43.407 43.568 multiply_cannon 2055 13.4 0.223 0.247 34.157 37.049 multiply_cannon_loop 2055 14.4 0.929 0.954 31.918 34.134 ot_mini 99 10.5 0.001 0.001 23.644 23.761 mp_waitall_1 176588 16.5 13.021 18.259 13.021 18.259 qs_ot_get_derivative 99 11.5 0.001 0.001 17.191 17.332 rebuild_ks_matrix 110 8.3 0.000 0.000 16.237 16.345 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 16.237 16.345 multiply_cannon_multrec 24660 15.4 6.951 8.981 13.173 15.154 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.296 14.390 multiply_cannon_metrocomm3 24660 15.4 0.068 0.071 8.296 13.235 init_scf_loop 11 6.9 0.001 0.005 13.105 13.107 qs_ot_get_p 110 10.4 0.001 0.001 12.014 12.152 prepare_preconditioner 11 7.9 0.000 0.001 10.771 10.788 make_preconditioner 11 8.9 0.000 0.000 10.771 10.788 make_full_inverse_cholesky 11 9.9 0.000 0.000 10.256 10.451 qs_ot_p2m_diag 48 11.0 0.029 0.044 8.941 8.978 sum_up_and_integrate 110 10.3 0.052 0.060 8.712 8.723 integrate_v_rspace 110 11.3 0.002 0.003 8.660 8.671 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.460 8.485 calculate_rho_elec 110 8.6 0.040 0.048 8.459 8.484 cp_dbcsr_syevd 48 12.0 0.003 0.003 8.227 8.230 multiply_cannon_sync_h2d 24660 15.4 6.471 7.826 6.471 7.826 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.822 7.353 apply_single 110 13.6 0.000 0.001 6.822 7.353 init_scf_run 11 5.9 0.000 0.001 7.186 7.186 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.186 7.186 make_m2s 4110 13.4 0.057 0.060 6.347 7.137 make_images 4110 14.4 0.401 0.447 6.235 7.021 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.007 7.018 cp_fm_diag_elpa_base 48 14.0 6.926 6.957 7.003 7.014 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.934 6.652 rs_pw_transfer 902 11.9 0.015 0.042 5.534 6.427 dbcsr_mm_accdrv_process 52282 16.1 4.592 5.498 6.079 6.392 density_rs2pw 110 9.6 0.004 0.005 5.422 6.260 ot_diis_step 99 11.5 0.010 0.010 6.234 6.234 cp_fm_cholesky_invert 11 10.9 5.848 5.860 5.848 5.860 pw_transfer 1331 11.6 0.068 0.074 5.286 5.465 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.177 5.360 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.096 5.160 mp_sum_l 6594 12.7 3.481 4.771 3.481 4.771 wfi_extrapolate 11 7.9 0.001 0.001 4.490 4.490 potential_pw2rs 110 12.3 0.008 0.009 4.352 4.394 fft3d_ps 1111 14.6 1.110 1.337 4.157 4.340 make_images_data 4110 15.4 0.047 0.051 3.420 4.338 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.942 3.900 fft_wrap_pw1pw2_140 451 13.1 0.513 0.532 3.404 3.583 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.514 3.517 calculate_dm_sparse 110 9.5 0.001 0.001 3.416 3.492 grid_integrate_task_list 110 12.3 3.121 3.386 3.121 3.386 cp_fm_cholesky_decompose 22 10.9 3.143 3.151 3.143 3.151 mp_alltoall_d11v 2046 13.8 2.596 3.132 2.596 3.132 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.000 3.033 mp_allgather_i34 2055 14.4 1.557 2.979 1.557 2.979 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.829 2.922 mp_alltoall_z22v 1111 16.6 2.628 2.736 2.628 2.736 calculate_first_density_matrix 1 7.0 0.002 0.015 2.565 2.569 grid_collocate_task_list 110 9.6 2.094 2.546 2.094 2.546 mp_waitany 10164 13.8 1.720 2.497 1.720 2.497 multiply_cannon_metrocomm4 22605 15.4 0.077 0.083 0.873 2.371 make_images_sizes 4110 15.4 0.005 0.005 1.564 2.157 mp_alltoall_i44 4110 16.4 1.559 2.152 1.559 2.152 mp_sum_d 3889 11.9 1.510 2.148 1.510 2.148 mp_irecv_dv 57340 16.2 0.748 2.141 0.748 2.141 jit_kernel_multiply 11 16.2 1.133 2.133 1.133 2.133 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.111 2.123 dbcsr_complete_redistribute 325 12.2 0.226 0.286 1.793 2.121 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.220 1.147 2.001 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.966 1.968 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=93.148000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.545455, yerr=6.344126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 660.340736E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.033 71.845 71.846 qs_mol_dyn_low 1 2.0 0.003 0.003 71.507 71.517 qs_forces 11 3.9 0.003 0.003 71.316 71.318 qs_energies 11 4.9 0.001 0.002 67.892 67.896 scf_env_do_scf 11 5.9 0.000 0.001 59.129 59.129 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 47.790 47.791 velocity_verlet 10 3.0 0.001 0.001 38.863 38.869 dbcsr_multiply_generic 2055 12.4 0.107 0.112 30.788 31.031 qs_scf_new_mos 99 7.5 0.001 0.001 29.990 30.096 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.989 30.096 ot_scf_mini 99 9.5 0.002 0.002 28.680 28.800 multiply_cannon 2055 13.4 0.215 0.226 22.768 24.111 multiply_cannon_loop 2055 14.4 0.618 0.632 21.223 22.213 ot_mini 99 10.5 0.001 0.001 15.289 15.403 rebuild_ks_matrix 110 8.3 0.000 0.000 14.050 14.183 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.050 14.182 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.394 12.513 mp_waitall_1 139946 16.5 8.634 11.757 8.634 11.757 init_scf_loop 11 6.9 0.000 0.000 11.285 11.286 qs_ot_get_derivative 99 11.5 0.001 0.001 10.769 10.889 multiply_cannon_multrec 16440 15.4 3.972 4.992 9.736 10.727 prepare_preconditioner 11 7.9 0.000 0.000 9.307 9.325 make_preconditioner 11 8.9 0.000 0.000 9.307 9.325 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.501 8.943 qs_ot_get_p 110 10.4 0.001 0.001 8.265 8.394 sum_up_and_integrate 110 10.3 0.060 0.062 7.900 7.916 integrate_v_rspace 110 11.3 0.003 0.003 7.840 7.856 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.546 7.563 calculate_rho_elec 110 8.6 0.059 0.060 7.545 7.562 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.283 7.380 init_scf_run 11 5.9 0.000 0.001 6.223 6.224 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.223 6.223 make_m2s 4110 13.4 0.050 0.051 5.625 6.160 qs_ot_p2m_diag 48 11.0 0.042 0.044 6.083 6.093 make_images 4110 14.4 0.394 0.513 5.508 6.042 dbcsr_mm_accdrv_process 34862 16.1 4.649 5.318 5.618 5.738 density_rs2pw 110 9.6 0.004 0.005 4.452 5.736 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.637 5.638 rs_pw_transfer 902 11.9 0.010 0.012 4.358 5.599 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.028 5.456 apply_single 110 13.6 0.000 0.000 5.028 5.456 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.851 4.859 cp_fm_diag_elpa_base 48 14.0 4.759 4.805 4.848 4.856 pw_transfer 1331 11.6 0.066 0.075 4.606 4.659 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.498 4.552 ot_diis_step 99 11.5 0.011 0.011 4.485 4.486 multiply_cannon_sync_h2d 16440 15.4 3.646 4.228 3.646 4.228 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.574 4.214 cp_fm_cholesky_invert 11 10.9 4.205 4.214 4.205 4.214 make_images_data 4110 15.4 0.043 0.047 3.098 3.963 hybrid_alltoall_any 4261 16.3 0.106 0.379 2.681 3.811 wfi_extrapolate 11 7.9 0.001 0.001 3.620 3.620 fft_wrap_pw1pw2_140 451 13.1 0.632 0.646 3.515 3.567 potential_pw2rs 110 12.3 0.011 0.012 3.443 3.491 grid_integrate_task_list 110 12.3 3.153 3.463 3.153 3.463 fft3d_ps 1111 14.6 1.094 1.104 3.325 3.378 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.022 3.088 mp_alltoall_d11v 2046 13.8 2.469 3.017 2.469 3.017 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.952 2.955 mp_waitany 17072 13.8 1.430 2.794 1.430 2.794 calculate_dm_sparse 110 9.5 0.001 0.001 2.620 2.653 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.511 2.566 grid_collocate_task_list 110 9.6 2.124 2.511 2.124 2.511 calculate_first_density_matrix 1 7.0 0.000 0.000 2.479 2.480 mp_sum_l 6594 12.7 1.782 2.452 1.782 2.452 dbcsr_complete_redistribute 325 12.2 0.309 0.340 1.891 2.419 cp_fm_cholesky_decompose 22 10.9 2.378 2.398 2.378 2.398 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.863 2.387 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.181 0.970 2.285 mp_irecv_dv 48980 15.7 0.796 2.263 0.796 2.263 jit_kernel_multiply 12 16.4 0.579 2.258 0.579 2.258 mp_alltoall_z22v 1111 16.6 2.020 2.070 2.020 2.070 mp_allgather_i34 2055 14.4 0.857 2.031 0.857 2.031 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.985 1.986 cp_fm_upper_to_full 70 14.2 1.457 1.945 1.457 1.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.855 1.870 make_images_sizes 4110 15.4 0.005 0.005 1.177 1.846 mp_alltoall_i44 4110 16.4 1.172 1.841 1.172 1.841 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.266 1.775 rs_gather_matrices 110 12.3 0.232 0.264 1.153 1.659 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.524 1.534 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.354 1.451 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.416 1.438 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=71.846000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.000000, yerr=7.942979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 732.385280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.059 75.532 75.534 qs_mol_dyn_low 1 2.0 0.003 0.003 75.175 75.186 qs_forces 11 3.9 0.003 0.003 74.973 74.974 qs_energies 11 4.9 0.007 0.032 71.377 71.387 scf_env_do_scf 11 5.9 0.001 0.001 61.906 61.908 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 48.154 48.155 velocity_verlet 10 3.0 0.001 0.001 41.689 41.696 dbcsr_multiply_generic 2055 12.4 0.113 0.118 31.773 32.012 qs_scf_new_mos 99 7.5 0.001 0.001 30.894 30.998 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.894 30.998 ot_scf_mini 99 9.5 0.003 0.003 29.194 29.303 multiply_cannon 2055 13.4 0.243 0.264 23.099 24.347 multiply_cannon_loop 2055 14.4 0.892 0.910 21.574 22.173 ot_mini 99 10.5 0.001 0.001 15.696 15.832 multiply_cannon_multrec 24660 15.4 4.244 6.909 12.896 14.236 init_scf_loop 11 6.9 0.001 0.005 13.692 13.695 rebuild_ks_matrix 110 8.3 0.000 0.000 13.536 13.636 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.536 13.636 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.974 12.062 prepare_preconditioner 11 7.9 0.000 0.001 11.777 11.797 make_preconditioner 11 8.9 0.001 0.002 11.777 11.797 qs_ot_get_derivative 99 11.5 0.001 0.001 11.466 11.585 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.932 11.424 dbcsr_mm_accdrv_process 52304 16.0 7.094 8.660 8.505 9.400 qs_ot_get_p 110 10.4 0.001 0.002 8.225 8.382 mp_waitall_1 121746 16.5 5.975 8.091 5.975 8.091 sum_up_and_integrate 110 10.3 0.068 0.071 7.585 7.599 integrate_v_rspace 110 11.3 0.003 0.003 7.517 7.531 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.355 7.365 calculate_rho_elec 110 8.6 0.078 0.082 7.355 7.365 make_m2s 4110 13.4 0.059 0.062 6.845 7.144 make_images 4110 14.4 0.577 0.698 6.703 7.003 init_scf_run 11 5.9 0.000 0.001 6.435 6.435 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.434 6.435 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.946 5.961 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.377 5.377 cp_fm_upper_to_full 70 14.2 3.461 5.017 3.461 5.017 density_rs2pw 110 9.6 0.004 0.004 4.108 4.623 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.599 4.608 cp_fm_diag_elpa_base 48 14.0 4.408 4.478 4.596 4.605 pw_transfer 1331 11.6 0.066 0.077 4.563 4.600 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.456 4.497 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.097 4.205 apply_single 110 13.6 0.000 0.000 4.096 4.205 ot_diis_step 99 11.5 0.011 0.011 4.186 4.186 rs_pw_transfer 902 11.9 0.010 0.011 3.708 4.175 dbcsr_complete_redistribute 325 12.2 0.413 0.466 2.994 4.161 make_images_data 4110 15.4 0.046 0.050 3.618 4.108 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.978 3.888 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.728 3.791 cp_fm_cholesky_invert 11 10.9 3.767 3.779 3.767 3.779 fft_wrap_pw1pw2_140 451 13.1 0.666 0.686 3.559 3.603 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.534 3.591 copy_fm_to_dbcsr 174 11.2 0.001 0.003 2.346 3.497 wfi_extrapolate 11 7.9 0.001 0.001 3.479 3.479 grid_integrate_task_list 110 12.3 3.232 3.433 3.232 3.433 multiply_cannon_sync_h2d 24660 15.4 3.140 3.262 3.140 3.262 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.477 3.254 fft3d_ps 1111 14.6 1.090 1.121 3.221 3.242 calculate_dm_sparse 110 9.5 0.001 0.001 3.127 3.156 potential_pw2rs 110 12.3 0.013 0.014 3.070 3.085 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.066 3.068 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.835 2.972 mp_alltoall_i22 605 13.7 1.734 2.925 1.734 2.925 calculate_first_density_matrix 1 7.0 0.001 0.006 2.831 2.833 mp_alltoall_d11v 2046 13.8 2.356 2.690 2.356 2.690 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.583 2.622 grid_collocate_task_list 110 9.6 2.221 2.503 2.221 2.503 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 2.369 2.371 cp_fm_cholesky_decompose 22 10.9 2.316 2.365 2.316 2.365 mp_alltoall_z22v 1111 16.6 1.914 1.979 1.914 1.979 jit_kernel_multiply 10 15.7 1.076 1.975 1.076 1.975 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.845 1.882 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.816 1.826 mp_waitany 13376 13.8 1.323 1.790 1.323 1.790 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.694 1.709 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.696 mp_sum_l 6594 12.7 1.147 1.665 1.147 1.665 make_images_sizes 4110 15.4 0.005 0.005 0.959 1.649 mp_alltoall_i44 4110 16.4 0.955 1.644 0.955 1.644 mp_allgather_i34 2055 14.4 0.621 1.624 0.621 1.624 acc_transpose_blocks 24660 15.4 0.104 0.106 1.592 1.615 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.844 1.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=75.534000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.818182, yerr=6.873112 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 836.792320E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 69.295 69.296 qs_mol_dyn_low 1 2.0 0.003 0.003 68.959 68.969 qs_forces 11 3.9 0.003 0.003 68.885 68.886 qs_energies 11 4.9 0.001 0.002 64.934 64.941 scf_env_do_scf 11 5.9 0.000 0.001 55.561 55.561 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.758 44.759 velocity_verlet 10 3.0 0.001 0.001 38.137 38.141 qs_scf_new_mos 99 7.5 0.001 0.001 27.129 27.208 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.128 27.207 dbcsr_multiply_generic 2055 12.4 0.106 0.111 26.514 26.809 ot_scf_mini 99 9.5 0.002 0.002 25.658 25.712 multiply_cannon 2055 13.4 0.249 0.264 18.946 20.845 multiply_cannon_loop 2055 14.4 0.325 0.337 17.302 18.005 rebuild_ks_matrix 110 8.3 0.000 0.000 13.521 13.545 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.521 13.544 ot_mini 99 10.5 0.001 0.001 12.892 12.907 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.970 11.989 init_scf_loop 11 6.9 0.000 0.000 10.738 10.740 mp_waitall_1 103326 16.6 8.405 10.230 8.405 10.230 qs_ot_get_derivative 99 11.5 0.001 0.001 8.954 9.007 prepare_preconditioner 11 7.9 0.000 0.000 8.875 8.878 make_preconditioner 11 8.9 0.000 0.000 8.875 8.878 multiply_cannon_multrec 8220 15.4 3.227 4.486 7.636 8.619 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.319 8.477 qs_ot_get_p 110 10.4 0.001 0.001 8.385 8.422 sum_up_and_integrate 110 10.3 0.079 0.081 7.774 7.787 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.745 7.762 calculate_rho_elec 110 8.6 0.115 0.115 7.745 7.762 integrate_v_rspace 110 11.3 0.003 0.003 7.694 7.706 qs_ot_p2m_diag 48 11.0 0.081 0.084 6.340 6.354 make_m2s 4110 13.4 0.039 0.041 5.699 6.019 init_scf_run 11 5.9 0.000 0.001 5.995 5.995 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.995 5.995 make_images 4110 14.4 0.638 0.691 5.567 5.887 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.866 5.867 dbcsr_mm_accdrv_process 17442 15.9 2.964 3.750 4.278 5.247 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.014 5.022 cp_fm_diag_elpa_base 48 14.0 4.919 4.973 5.011 5.019 pw_transfer 1331 11.6 0.066 0.072 4.956 4.989 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.848 4.884 density_rs2pw 110 9.6 0.004 0.004 4.226 4.605 cp_fm_cholesky_invert 11 10.9 4.545 4.551 4.545 4.551 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.775 4.167 make_images_data 4110 15.4 0.038 0.043 3.143 4.004 fft_wrap_pw1pw2_140 451 13.1 0.830 0.844 3.962 4.002 ot_diis_step 99 11.5 0.012 0.012 3.872 3.872 rs_pw_transfer 902 11.9 0.010 0.010 3.486 3.823 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.781 3.803 apply_single 110 13.6 0.000 0.000 3.781 3.802 hybrid_alltoall_any 4261 16.3 0.201 0.866 2.885 3.727 grid_integrate_task_list 110 12.3 3.406 3.598 3.406 3.598 fft3d_ps 1111 14.6 1.148 1.168 3.396 3.436 wfi_extrapolate 11 7.9 0.001 0.001 3.415 3.415 potential_pw2rs 110 12.3 0.015 0.016 3.043 3.077 multiply_cannon_sync_h2d 8220 15.4 2.893 3.014 2.893 3.014 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 2.108 2.922 calculate_dm_sparse 110 9.5 0.001 0.001 2.791 2.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.762 2.765 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.740 2.760 cp_fm_cholesky_decompose 22 10.9 2.702 2.716 2.702 2.716 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.694 2.695 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.657 2.680 mp_alltoall_d11v 2046 13.8 2.430 2.677 2.430 2.677 grid_collocate_task_list 110 9.6 2.324 2.590 2.324 2.590 calculate_first_density_matrix 1 7.0 0.000 0.000 2.451 2.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.312 2.321 mp_allgather_i34 2055 14.4 0.819 2.138 0.819 2.138 mp_alltoall_z22v 1111 16.6 2.007 2.047 2.007 2.047 dbcsr_complete_redistribute 325 12.2 0.545 0.597 1.897 2.017 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.766 1.988 make_images_sizes 4110 15.4 0.005 0.005 1.136 1.987 mp_alltoall_i44 4110 16.4 1.131 1.983 1.131 1.983 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.887 1.891 mp_waitany 9240 13.8 1.317 1.756 1.317 1.756 mp_sum_l 6594 12.7 1.173 1.660 1.173 1.660 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.532 1.656 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.565 1.574 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.371 1.389 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=69.296000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=791.909091, yerr=9.829959 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 1.361658E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 100.799 100.800 qs_mol_dyn_low 1 2.0 0.003 0.003 100.432 100.442 qs_forces 11 3.9 0.003 0.003 100.349 100.350 qs_energies 11 4.9 0.002 0.002 95.919 95.921 scf_env_do_scf 11 5.9 0.001 0.001 85.228 85.229 velocity_verlet 10 3.0 0.001 0.001 62.395 62.402 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 53.281 53.283 init_scf_loop 11 6.9 0.000 0.000 31.861 31.866 qs_scf_new_mos 99 7.5 0.001 0.001 31.803 31.866 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.802 31.865 dbcsr_multiply_generic 2055 12.4 0.118 0.123 31.273 31.357 ot_scf_mini 99 9.5 0.002 0.002 29.910 29.952 prepare_preconditioner 11 7.9 0.000 0.000 29.607 29.618 make_preconditioner 11 8.9 0.000 0.000 29.607 29.618 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.467 29.024 multiply_cannon 2055 13.4 0.346 0.362 22.072 23.067 multiply_cannon_loop 2055 14.4 0.342 0.346 20.072 20.406 cp_fm_upper_to_full 70 14.2 13.066 18.948 13.066 18.948 rebuild_ks_matrix 110 8.3 0.000 0.000 16.020 16.067 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 16.020 16.067 ot_mini 99 10.5 0.001 0.001 15.560 15.604 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.516 14.558 dbcsr_complete_redistribute 325 12.2 1.013 1.039 8.035 11.479 qs_ot_get_derivative 99 11.5 0.001 0.001 10.817 10.875 mp_waitall_1 84994 16.7 9.655 10.741 9.655 10.741 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.848 10.280 multiply_cannon_multrec 8220 15.4 4.369 4.514 9.600 9.712 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.508 9.550 calculate_rho_elec 110 8.6 0.226 0.227 9.508 9.550 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.125 9.512 mp_alltoall_i22 605 13.7 5.736 9.187 5.736 9.187 qs_ot_get_p 110 10.4 0.001 0.001 8.774 8.827 sum_up_and_integrate 110 10.3 0.150 0.151 8.751 8.768 integrate_v_rspace 110 11.3 0.003 0.003 8.601 8.618 make_m2s 4110 13.4 0.043 0.044 7.088 7.689 make_images 4110 14.4 0.882 0.928 6.895 7.496 cp_fm_cholesky_invert 11 10.9 6.985 6.991 6.985 6.991 qs_ot_p2m_diag 48 11.0 0.151 0.156 6.723 6.729 pw_transfer 1331 11.6 0.075 0.076 6.616 6.622 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.498 6.506 init_scf_run 11 5.9 0.000 0.001 6.286 6.287 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.286 6.286 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.178 6.179 fft_wrap_pw1pw2_140 451 13.1 1.337 1.346 5.414 5.424 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.994 5.346 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.819 5.334 apply_single 110 13.6 0.000 0.000 4.819 5.334 dbcsr_mm_accdrv_process 11614 15.7 3.342 3.652 5.089 5.311 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.293 5.293 cp_fm_diag_elpa_base 48 14.0 4.660 4.951 5.290 5.290 density_rs2pw 110 9.6 0.004 0.004 5.206 5.244 make_images_data 4110 15.4 0.041 0.044 4.083 5.166 hybrid_alltoall_any 4261 16.3 0.256 0.552 3.896 4.907 ot_diis_step 99 11.5 0.015 0.016 4.696 4.697 fft3d_ps 1111 14.6 1.312 1.322 4.429 4.441 multiply_cannon_sync_h2d 8220 15.4 3.949 3.952 3.949 3.952 wfi_extrapolate 11 7.9 0.001 0.001 3.827 3.827 grid_integrate_task_list 110 12.3 3.713 3.777 3.713 3.777 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.273 3.757 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.690 3.691 rs_pw_transfer 902 11.9 0.010 0.011 3.587 3.616 potential_pw2rs 110 12.3 0.021 0.022 3.520 3.526 calculate_dm_sparse 110 9.5 0.001 0.001 3.287 3.313 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.155 3.192 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.995 2.997 mp_alltoall_d11v 2046 13.8 2.886 2.988 2.886 2.988 cp_fm_cholesky_decompose 22 10.9 2.934 2.953 2.934 2.953 mp_alltoall_z22v 1111 16.6 2.863 2.876 2.863 2.876 grid_collocate_task_list 110 9.6 2.701 2.737 2.701 2.737 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.404 2.421 calculate_first_density_matrix 1 7.0 0.000 0.000 2.302 2.303 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.211 2.268 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.184 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.113 2.117 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.010 2.039 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=100.800000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1226.090909, yerr=50.057157 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.725504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 1802496. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 212.551 212.552 qs_mol_dyn_low 1 2.0 0.003 0.003 212.085 212.099 qs_forces 11 3.9 0.005 0.005 211.996 211.998 qs_energies 11 4.9 0.002 0.003 206.247 206.266 scf_env_do_scf 11 5.9 0.001 0.002 189.516 189.519 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 168.137 168.139 qs_scf_new_mos 117 7.6 0.001 0.001 126.672 126.941 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.671 126.940 velocity_verlet 10 3.0 0.001 0.001 126.913 126.914 dbcsr_multiply_generic 2507 12.6 0.177 0.180 125.251 125.812 ot_scf_mini 117 9.6 0.003 0.003 120.032 120.304 multiply_cannon 2507 13.6 0.239 0.246 101.213 102.669 multiply_cannon_loop 2507 14.6 2.090 2.134 98.939 100.204 ot_mini 117 10.6 0.001 0.001 66.394 66.655 multiply_cannon_multrec 60168 15.6 33.228 34.699 41.544 43.112 qs_ot_get_derivative 117 11.6 0.001 0.001 41.623 41.874 rebuild_ks_matrix 128 8.3 0.001 0.001 34.477 34.718 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.476 34.717 mp_waitall_1 267128 16.5 28.984 32.542 28.984 32.542 qs_ot_get_p 128 10.4 0.001 0.001 31.603 31.863 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.918 31.160 multiply_cannon_sync_h2d 60168 15.6 27.465 28.879 27.465 28.879 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.330 25.125 apply_single 128 13.6 0.001 0.001 24.330 25.125 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.858 24.915 ot_diis_step 117 11.6 0.008 0.008 24.540 24.541 cp_dbcsr_syevd 83 12.4 0.005 0.005 22.299 22.300 init_scf_loop 11 6.9 0.000 0.001 21.290 21.291 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.956 20.195 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.117 19.153 cp_fm_diag_elpa_base 83 14.4 19.034 19.071 19.112 19.146 multiply_cannon_metrocomm3 60168 15.6 0.111 0.115 15.789 17.751 prepare_preconditioner 11 7.9 0.000 0.000 16.605 16.646 make_preconditioner 11 8.9 0.000 0.000 16.605 16.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.798 16.000 sum_up_and_integrate 128 10.3 0.090 0.109 14.812 14.830 integrate_v_rspace 128 11.3 0.004 0.005 14.722 14.743 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.621 14.741 calculate_rho_elec 128 8.7 0.046 0.065 14.620 14.740 make_m2s 5014 13.6 0.105 0.112 14.255 14.676 make_images 5014 14.6 0.404 0.423 14.076 14.505 init_scf_run 11 5.9 0.000 0.001 12.574 12.574 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.573 12.574 density_rs2pw 128 9.7 0.006 0.007 7.806 11.092 rs_pw_transfer 1046 11.9 0.017 0.018 6.512 9.865 mp_sum_l 7950 12.9 8.522 9.504 8.522 9.504 cp_fm_cholesky_invert 11 10.9 9.473 9.481 9.473 9.481 wfi_extrapolate 11 7.9 0.001 0.001 9.260 9.260 calculate_dm_sparse 128 9.5 0.001 0.001 8.516 8.597 pw_transfer 1547 11.6 0.075 0.093 8.266 8.458 multiply_cannon_metrocomm1 60168 15.6 0.093 0.099 6.122 8.457 dbcsr_mm_accdrv_process 124484 16.2 3.202 3.329 7.879 8.392 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 8.062 8.247 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.012 8.182 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.924 8.005 make_images_data 5014 15.6 0.067 0.073 7.021 7.832 grid_integrate_task_list 128 12.3 7.065 7.645 7.065 7.645 hybrid_alltoall_any 5200 16.5 0.291 2.276 6.130 7.281 fft_wrap_pw1pw2_140 523 13.2 1.275 1.318 6.855 7.043 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.679 6.691 fft3d_ps 1291 14.7 2.140 2.776 5.824 6.154 mp_waitany 16020 13.9 2.698 6.095 2.698 6.095 grid_collocate_task_list 128 9.7 4.688 5.939 4.688 5.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.759 5.873 mp_alltoall_d11v 2415 14.1 4.570 5.752 4.570 5.752 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.293 2.151 5.450 potential_pw2rs 128 12.3 0.009 0.011 5.078 5.129 cp_fm_cholesky_decompose 22 10.9 4.911 4.924 4.911 4.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=212.552000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=6.707588 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.775296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58290. MP_Allreduce 11071 960. MP_Sync 86 MP_Alltoall 1955 5616340. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.030 208.952 208.953 qs_mol_dyn_low 1 2.0 0.003 0.005 208.522 208.536 qs_forces 11 3.9 0.004 0.005 208.365 208.367 qs_energies 11 4.9 0.002 0.006 201.212 201.222 scf_env_do_scf 11 5.9 0.001 0.001 183.998 184.008 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 146.759 146.761 velocity_verlet 10 3.0 0.001 0.002 129.779 129.782 qs_scf_new_mos 116 7.6 0.001 0.001 103.151 103.700 qs_scf_loop_do_ot 116 8.6 0.001 0.001 103.151 103.699 dbcsr_multiply_generic 2485 12.5 0.187 0.193 98.851 99.959 ot_scf_mini 116 9.6 0.004 0.004 98.313 98.922 multiply_cannon 2485 13.5 0.472 0.528 77.396 81.199 multiply_cannon_loop 2485 14.5 1.235 1.274 73.908 76.323 ot_mini 116 10.6 0.001 0.001 51.051 51.594 mp_waitall_1 212858 16.6 26.310 39.293 26.310 39.293 init_scf_loop 11 6.9 0.000 0.000 37.137 37.139 multiply_cannon_multrec 29820 15.5 22.087 26.660 31.654 36.482 rebuild_ks_matrix 127 8.3 0.001 0.001 35.462 35.800 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 35.461 35.799 prepare_preconditioner 11 7.9 0.000 0.000 32.464 32.542 make_preconditioner 11 8.9 0.000 0.000 32.464 32.542 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.829 32.145 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.990 31.605 qs_ot_get_derivative 116 11.6 0.001 0.002 29.257 29.860 qs_ot_get_p 127 10.4 0.001 0.001 28.831 29.414 multiply_cannon_metrocomm3 29820 15.5 0.090 0.094 15.334 27.634 qs_ot_p2m_diag 82 11.4 0.185 0.213 23.432 23.471 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.866 23.020 apply_single 127 13.6 0.001 0.001 21.865 23.020 cp_dbcsr_syevd 82 12.4 0.006 0.006 22.130 22.134 multiply_cannon_sync_h2d 29820 15.5 19.398 21.825 19.398 21.825 ot_diis_step 116 11.6 0.014 0.015 21.619 21.623 cp_fm_cholesky_invert 11 10.9 18.509 18.522 18.509 18.522 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.257 18.286 cp_fm_diag_elpa_base 82 14.4 17.948 18.075 18.250 18.278 make_m2s 4970 13.5 0.090 0.095 15.702 17.232 make_images 4970 14.5 1.160 1.349 15.494 17.021 sum_up_and_integrate 127 10.3 0.116 0.132 16.706 16.737 integrate_v_rspace 127 11.3 0.004 0.004 16.590 16.626 qs_rho_update_rho_low 127 7.7 0.001 0.002 16.146 16.185 calculate_rho_elec 127 8.7 0.088 0.105 16.145 16.184 init_scf_run 11 5.9 0.000 0.001 12.273 12.274 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.272 12.274 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.646 12.088 density_rs2pw 127 9.7 0.007 0.007 9.403 11.795 make_images_data 4970 15.5 0.064 0.073 9.282 11.169 rs_pw_transfer 1038 11.9 0.014 0.017 8.753 11.028 multiply_cannon_metrocomm4 27335 15.5 0.101 0.115 3.617 10.269 hybrid_alltoall_any 5155 16.4 0.341 1.495 7.821 10.241 mp_irecv_dv 68888 16.3 3.422 9.884 3.422 9.884 pw_transfer 1535 11.6 0.087 0.094 9.720 9.839 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 9.496 9.615 dbcsr_mm_accdrv_process 61726 16.2 4.429 5.122 9.021 9.603 wfi_extrapolate 11 7.9 0.001 0.001 9.109 9.111 cp_fm_cholesky_decompose 22 10.9 8.870 8.948 8.870 8.948 fft_wrap_pw1pw2_140 519 13.2 1.321 1.344 7.921 8.063 grid_integrate_task_list 127 12.3 7.071 7.748 7.071 7.748 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.457 7.216 potential_pw2rs 127 12.3 0.016 0.019 6.869 6.953 fft3d_ps 1281 14.7 2.793 2.966 6.825 6.922 calculate_dm_sparse 127 9.5 0.001 0.001 6.509 6.660 mp_alltoall_d11v 2401 14.1 4.723 6.286 4.723 6.286 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.266 6.276 grid_collocate_task_list 127 9.7 4.740 5.919 4.740 5.919 mp_sum_l 7884 12.9 3.988 5.916 3.988 5.916 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.443 5.516 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.311 5.429 mp_waitany 11660 13.9 2.750 5.252 2.750 5.252 mp_allgather_i34 2485 14.5 2.053 4.972 2.053 4.972 dbcsr_complete_redistribute 393 12.7 0.773 0.855 3.654 4.623 rs_pw_transfer_RS2PW_140 138 11.5 0.349 0.372 2.155 4.507 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=208.953000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.909091, yerr=2.274545 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 948.596736E+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 3992 58357. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.031 212.505 212.506 qs_mol_dyn_low 1 2.0 0.003 0.003 212.049 212.063 qs_forces 11 3.9 0.004 0.004 211.935 211.941 qs_energies 11 4.9 0.001 0.002 204.968 204.986 scf_env_do_scf 11 5.9 0.001 0.001 188.160 188.160 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 143.762 143.764 velocity_verlet 10 3.0 0.001 0.002 134.163 134.166 qs_scf_new_mos 117 7.6 0.001 0.001 101.251 101.711 qs_scf_loop_do_ot 117 8.6 0.001 0.001 101.251 101.711 ot_scf_mini 117 9.6 0.003 0.004 96.721 97.236 dbcsr_multiply_generic 2507 12.6 0.179 0.185 88.861 90.611 multiply_cannon 2507 13.6 0.499 0.524 64.719 68.623 multiply_cannon_loop 2507 14.6 0.861 0.892 60.715 65.575 ot_mini 117 10.6 0.001 0.001 48.314 48.879 init_scf_loop 11 6.9 0.000 0.000 44.267 44.269 mp_waitall_1 170520 16.6 30.253 42.724 30.253 42.724 prepare_preconditioner 11 7.9 0.000 0.000 39.940 39.998 make_preconditioner 11 8.9 0.000 0.000 39.940 39.998 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.335 38.897 rebuild_ks_matrix 128 8.3 0.001 0.001 33.331 33.949 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.330 33.948 qs_ot_get_p 128 10.4 0.001 0.001 32.776 33.463 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.028 30.594 multiply_cannon_metrocomm3 20056 15.6 0.058 0.063 17.590 29.748 qs_ot_get_derivative 117 11.6 0.001 0.002 27.750 28.259 qs_ot_p2m_diag 83 11.4 0.266 0.273 27.212 27.275 cp_dbcsr_syevd 83 12.4 0.005 0.006 25.776 25.782 multiply_cannon_multrec 20056 15.6 13.179 16.598 21.803 25.249 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.325 21.651 apply_single 128 13.6 0.001 0.001 20.325 21.651 cp_fm_diag_elpa 83 13.4 0.000 0.001 21.283 21.292 cp_fm_diag_elpa_base 83 14.4 20.781 20.989 21.275 21.285 ot_diis_step 117 11.6 0.018 0.021 20.341 20.342 cp_fm_cholesky_invert 11 10.9 20.162 20.176 20.162 20.176 make_m2s 5014 13.6 0.080 0.083 17.107 18.292 make_images 5014 14.6 1.191 1.284 16.876 18.063 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.627 16.724 calculate_rho_elec 128 8.7 0.132 0.148 16.627 16.724 sum_up_and_integrate 128 10.3 0.134 0.147 16.675 16.702 integrate_v_rspace 128 11.3 0.004 0.004 16.541 16.570 multiply_cannon_sync_h2d 20056 15.6 13.978 15.858 13.978 15.858 make_images_data 5014 15.6 0.059 0.067 10.466 12.276 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.544 11.936 init_scf_run 11 5.9 0.000 0.001 11.536 11.536 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.536 11.536 density_rs2pw 128 9.7 0.006 0.007 9.158 11.175 hybrid_alltoall_any 5200 16.5 0.437 2.006 8.906 10.890 cp_fm_cholesky_decompose 22 10.9 10.217 10.236 10.217 10.236 pw_transfer 1547 11.6 0.088 0.104 9.956 10.123 rs_pw_transfer 1046 11.9 0.013 0.016 8.005 10.115 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.730 9.901 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.475 9.753 mp_irecv_dv 50230 16.2 3.357 9.498 3.357 9.498 wfi_extrapolate 11 7.9 0.001 0.001 8.424 8.425 dbcsr_mm_accdrv_process 41502 16.2 4.592 5.253 8.135 8.248 fft_wrap_pw1pw2_140 523 13.2 1.412 1.445 8.029 8.214 cp_fm_upper_to_full 105 14.8 6.077 7.819 6.077 7.819 grid_integrate_task_list 128 12.3 7.222 7.753 7.222 7.753 mp_sum_l 7950 12.9 5.239 7.592 5.239 7.592 dbcsr_complete_redistribute 395 12.7 1.172 1.203 5.309 7.385 fft3d_ps 1291 14.7 2.742 2.974 6.851 6.976 potential_pw2rs 128 12.3 0.021 0.024 6.580 6.644 mp_alltoall_d11v 2415 14.1 5.251 6.491 5.251 6.491 calculate_dm_sparse 128 9.5 0.001 0.001 6.155 6.232 mp_allgather_i34 2507 14.6 2.504 6.182 2.504 6.182 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.338 6.075 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.819 5.893 grid_collocate_task_list 128 9.7 4.992 5.765 4.992 5.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.652 5.661 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.704 4.754 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.592 4.752 mp_waitany 11748 13.9 2.716 4.748 2.716 4.748 mp_alltoall_i22 716 14.1 2.348 4.627 2.348 4.627 transfer_fm_to_dbcsr 11 9.9 0.020 0.027 2.586 4.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=212.506000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=900.636364, yerr=7.713892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/18/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 4.387242E+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 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.140679E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57632. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509627. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.036 232.168 232.168 qs_mol_dyn_low 1 2.0 0.003 0.003 231.736 231.751 qs_forces 11 3.9 0.004 0.005 231.606 231.608 qs_energies 11 4.9 0.002 0.002 223.997 224.008 scf_env_do_scf 11 5.9 0.001 0.013 205.421 205.428 velocity_verlet 10 3.0 0.002 0.008 155.492 155.495 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 144.666 144.669 qs_scf_new_mos 118 7.6 0.001 0.001 102.868 103.215 qs_scf_loop_do_ot 118 8.6 0.001 0.001 102.867 103.215 ot_scf_mini 118 9.6 0.003 0.004 97.981 98.273 dbcsr_multiply_generic 2529 12.6 0.186 0.194 86.748 87.586 multiply_cannon 2529 13.6 0.557 0.605 58.233 61.125 init_scf_loop 11 6.9 0.000 0.000 60.579 60.584 prepare_preconditioner 11 7.9 0.000 0.000 56.064 56.088 make_preconditioner 11 8.9 0.000 0.000 56.064 56.088 multiply_cannon_loop 2529 14.6 1.196 1.220 53.052 55.224 make_full_inverse_cholesky 11 9.9 0.011 0.023 49.284 54.629 ot_mini 118 10.6 0.001 0.001 48.146 48.470 qs_ot_get_p 129 10.4 0.001 0.001 35.040 35.379 mp_waitall_1 149172 16.7 21.969 34.206 21.969 34.206 rebuild_ks_matrix 129 8.3 0.001 0.001 32.435 32.747 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.434 32.747 multiply_cannon_multrec 30348 15.6 13.814 18.851 26.199 31.104 qs_ot_p2m_diag 84 11.4 0.347 0.395 29.782 29.843 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.192 29.474 cp_dbcsr_syevd 84 12.4 0.005 0.006 28.223 28.228 qs_ot_get_derivative 118 11.6 0.001 0.002 27.250 27.539 make_m2s 5058 13.6 0.097 0.100 23.183 25.296 make_images 5058 14.6 1.973 2.289 22.874 24.985 cp_fm_cholesky_invert 11 10.9 24.449 24.471 24.449 24.471 cp_fm_diag_elpa 84 13.4 0.000 0.001 23.875 23.887 cp_fm_diag_elpa_base 84 14.4 22.726 23.124 23.866 23.877 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 20.152 20.746 apply_single 129 13.6 0.001 0.001 20.152 20.746 ot_diis_step 118 11.6 0.018 0.021 20.730 20.733 cp_fm_upper_to_full 106 14.8 12.182 18.022 12.182 18.022 multiply_cannon_metrocomm3 30348 15.6 0.048 0.051 7.850 17.709 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.128 17.157 calculate_rho_elec 129 8.7 0.177 0.194 17.127 17.156 sum_up_and_integrate 129 10.3 0.141 0.153 16.707 16.731 integrate_v_rspace 129 11.3 0.004 0.004 16.565 16.595 make_images_data 5058 15.6 0.062 0.068 12.885 15.903 dbcsr_complete_redistribute 397 12.7 1.521 1.667 10.082 14.112 hybrid_alltoall_any 5245 16.5 0.534 2.225 11.325 13.857 multiply_cannon_sync_h2d 30348 15.6 11.705 12.955 11.705 12.955 dbcsr_mm_accdrv_process 62780 16.2 7.591 8.690 11.957 12.486 init_scf_run 11 5.9 0.000 0.001 12.387 12.388 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.387 12.388 copy_fm_to_dbcsr 210 11.7 0.001 0.002 8.329 12.351 cp_fm_cholesky_decompose 22 10.9 12.220 12.320 12.220 12.320 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.525 11.744 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.757 10.712 pw_transfer 1559 11.6 0.088 0.102 10.487 10.578 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 10.257 10.354 density_rs2pw 129 9.7 0.006 0.006 8.842 10.250 mp_alltoall_i22 720 14.1 6.081 10.121 6.081 10.121 wfi_extrapolate 11 7.9 0.001 0.001 8.909 8.909 fft_wrap_pw1pw2_140 527 13.2 1.569 1.611 8.734 8.833 rs_pw_transfer 1054 12.0 0.014 0.021 7.024 8.523 grid_integrate_task_list 129 12.3 7.524 8.007 7.524 8.007 multiply_cannon_metrocomm4 25290 15.6 0.080 0.093 2.858 7.810 mp_irecv_dv 76751 16.2 2.713 7.536 2.713 7.536 mp_alltoall_d11v 2429 14.1 6.580 7.531 6.580 7.531 fft3d_ps 1301 14.7 2.864 2.955 7.158 7.215 calculate_dm_sparse 129 9.5 0.001 0.001 6.750 6.841 potential_pw2rs 129 12.3 0.023 0.024 6.150 6.184 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.903 5.956 grid_collocate_task_list 129 9.7 5.195 5.883 5.195 5.883 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.216 5.295 mp_allgather_i34 2529 14.6 3.257 4.824 3.257 4.824 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.718 4.786 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.591 4.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=232.168000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1077.636364, yerr=17.546161 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.527034E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108700816 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828148. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.032 198.966 198.967 qs_mol_dyn_low 1 2.0 0.003 0.003 198.450 198.464 qs_forces 11 3.9 0.004 0.004 198.332 198.337 qs_energies 11 4.9 0.002 0.002 190.443 190.454 scf_env_do_scf 11 5.9 0.001 0.002 171.784 171.788 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 129.489 129.491 velocity_verlet 10 3.0 0.001 0.001 128.045 128.050 qs_scf_new_mos 116 7.6 0.001 0.001 87.875 87.983 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.874 87.982 ot_scf_mini 116 9.6 0.003 0.004 83.283 83.356 dbcsr_multiply_generic 2485 12.5 0.179 0.190 77.922 78.443 multiply_cannon 2485 13.5 0.584 0.626 55.533 61.173 multiply_cannon_loop 2485 14.5 0.442 0.454 50.609 51.445 ot_mini 116 10.6 0.001 0.001 42.311 42.381 init_scf_loop 11 6.9 0.000 0.000 42.115 42.118 prepare_preconditioner 11 7.9 0.000 0.000 37.929 37.951 make_preconditioner 11 8.9 0.000 0.000 37.929 37.951 make_full_inverse_cholesky 11 9.9 0.016 0.025 35.622 36.031 mp_waitall_1 124680 16.7 28.672 35.919 28.672 35.919 rebuild_ks_matrix 127 8.3 0.001 0.001 31.659 31.772 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 31.659 31.772 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.666 28.770 qs_ot_get_p 127 10.4 0.001 0.001 27.780 27.910 qs_ot_p2m_diag 82 11.4 0.489 0.495 23.197 23.227 qs_ot_get_derivative 116 11.6 0.002 0.002 22.434 22.514 cp_dbcsr_syevd 82 12.4 0.005 0.005 21.829 21.834 cp_fm_cholesky_invert 11 10.9 21.623 21.630 21.623 21.630 multiply_cannon_multrec 9940 15.5 10.271 14.101 17.765 20.888 make_m2s 4970 13.5 0.067 0.073 18.296 20.659 make_images 4970 14.5 2.292 2.642 17.989 20.345 ot_diis_step 116 11.6 0.020 0.020 19.806 19.806 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.405 19.726 apply_single 127 13.6 0.001 0.001 19.405 19.725 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 11.721 18.706 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.986 17.995 cp_fm_diag_elpa_base 82 14.4 17.658 17.809 17.974 17.983 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.142 17.184 calculate_rho_elec 127 8.7 0.257 0.267 17.142 17.184 sum_up_and_integrate 127 10.3 0.178 0.189 16.465 16.508 integrate_v_rspace 127 11.3 0.004 0.005 16.286 16.339 make_images_data 4970 15.5 0.050 0.060 11.029 13.798 hybrid_alltoall_any 5155 16.4 0.837 3.815 10.556 13.404 multiply_cannon_sync_h2d 9940 15.5 11.478 11.943 11.478 11.943 init_scf_run 11 5.9 0.000 0.001 11.587 11.587 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.587 11.587 pw_transfer 1535 11.6 0.086 0.094 10.740 10.802 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.517 10.583 cp_fm_cholesky_decompose 22 10.9 10.325 10.446 10.325 10.446 density_rs2pw 127 9.7 0.006 0.006 8.355 9.256 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 5.985 9.252 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.135 9.187 fft_wrap_pw1pw2_140 519 13.2 1.897 1.930 8.981 9.046 wfi_extrapolate 11 7.9 0.001 0.001 8.331 8.331 grid_integrate_task_list 127 12.3 7.722 8.261 7.722 8.261 mp_allgather_i34 2485 14.5 3.115 8.203 3.115 8.203 dbcsr_mm_accdrv_process 20590 16.1 2.631 3.452 7.132 7.807 mp_alltoall_d11v 2401 14.1 6.241 7.646 6.241 7.646 fft3d_ps 1281 14.7 2.742 2.810 7.018 7.080 rs_pw_transfer 1038 11.9 0.012 0.013 5.716 6.685 calculate_dm_sparse 127 9.5 0.001 0.001 6.480 6.578 dbcsr_complete_redistribute 393 12.7 2.111 2.174 5.870 6.386 grid_collocate_task_list 127 9.7 5.425 6.100 5.425 6.100 potential_pw2rs 127 12.3 0.027 0.027 5.582 5.637 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.455 5.473 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.324 5.325 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.861 4.280 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.209 4.239 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.813 4.220 mp_irecv_dv 28618 15.9 1.830 4.210 1.830 4.210 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.009 4.035 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.911 4.020 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=198.967000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1430.545455, yerr=50.046094 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/20/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 11.696234E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.115741E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 58886. MP_Allreduce 11127 1502. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.044 327.116 327.116 qs_mol_dyn_low 1 2.0 0.003 0.003 326.465 326.478 qs_forces 11 3.9 0.004 0.004 326.145 326.148 qs_energies 11 4.9 0.002 0.002 316.291 316.302 scf_env_do_scf 11 5.9 0.001 0.002 291.601 291.610 velocity_verlet 10 3.0 0.001 0.001 233.665 233.701 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 157.003 157.006 init_scf_loop 11 6.9 0.000 0.000 134.303 134.309 prepare_preconditioner 11 7.9 0.000 0.000 129.019 129.053 make_preconditioner 11 8.9 0.000 0.000 129.019 129.053 make_full_inverse_cholesky 11 9.9 0.035 0.039 104.423 126.010 qs_scf_new_mos 117 7.6 0.001 0.001 101.911 102.068 qs_scf_loop_do_ot 117 8.6 0.001 0.001 101.910 102.067 ot_scf_mini 117 9.6 0.004 0.004 96.984 97.040 dbcsr_multiply_generic 2507 12.6 0.210 0.221 88.018 88.595 cp_fm_upper_to_full 105 14.8 54.060 77.648 54.060 77.648 multiply_cannon 2507 13.6 0.707 0.762 60.595 61.271 multiply_cannon_loop 2507 14.6 0.472 0.479 56.704 58.093 ot_mini 117 10.6 0.001 0.001 47.183 47.240 dbcsr_complete_redistribute 395 12.7 3.957 4.013 30.863 44.130 rebuild_ks_matrix 128 8.3 0.001 0.001 40.969 41.028 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 40.968 41.028 copy_fm_to_dbcsr 209 11.7 0.001 0.002 27.181 40.424 cp_fm_cholesky_invert 11 10.9 38.528 38.536 38.528 38.536 qs_ks_update_qs_env 128 7.6 0.001 0.001 37.585 37.645 transfer_fm_to_dbcsr 11 9.9 0.030 0.033 24.550 37.524 mp_alltoall_i22 716 14.1 22.277 35.570 22.277 35.570 mp_waitall_1 103674 16.8 30.978 35.472 30.978 35.472 qs_ot_get_p 128 10.4 0.001 0.001 33.789 33.840 qs_ot_p2m_diag 83 11.4 0.878 0.884 29.038 29.064 cp_dbcsr_syevd 83 12.4 0.006 0.006 27.179 27.183 qs_ot_get_derivative 117 11.6 0.002 0.002 26.627 26.684 cp_fm_diag_elpa 83 13.4 0.000 0.000 23.372 23.373 cp_fm_diag_elpa_base 83 14.4 18.751 20.581 23.366 23.366 make_m2s 5014 13.6 0.077 0.080 22.330 23.313 qs_rho_update_rho_low 128 7.7 0.001 0.001 22.963 23.001 calculate_rho_elec 128 8.7 0.481 0.482 22.962 23.000 make_images 5014 14.6 3.767 3.876 21.848 22.829 sum_up_and_integrate 128 10.3 0.321 0.323 21.543 21.625 integrate_v_rspace 128 11.3 0.004 0.004 21.222 21.304 ot_diis_step 117 11.6 0.022 0.024 20.512 20.512 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 19.267 20.510 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.632 19.899 apply_single 128 13.6 0.001 0.001 19.632 19.899 multiply_cannon_multrec 10028 15.6 10.525 12.252 18.257 18.404 multiply_cannon_sync_h2d 10028 15.6 15.656 15.665 15.656 15.665 make_images_data 5014 15.6 0.057 0.062 12.316 14.610 init_scf_run 11 5.9 0.000 0.001 14.378 14.378 scf_env_initial_rho_setup 11 6.9 0.000 0.000 14.378 14.378 hybrid_alltoall_any 5200 16.5 1.301 3.033 12.123 14.346 pw_transfer 1547 11.6 0.093 0.094 13.722 13.727 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 13.486 13.492 cp_fm_cholesky_decompose 22 10.9 11.758 11.816 11.758 11.816 fft_wrap_pw1pw2_140 523 13.2 3.071 3.112 11.702 11.718 mp_alltoall_d11v 2415 14.1 10.445 11.438 10.445 11.438 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.571 10.619 wfi_extrapolate 11 7.9 0.001 0.001 10.087 10.087 density_rs2pw 128 9.7 0.005 0.006 9.780 9.928 dbcsr_mm_accdrv_process 20762 16.1 3.754 5.724 7.492 9.366 grid_integrate_task_list 128 12.3 8.601 8.774 8.601 8.774 fft3d_ps 1291 14.7 2.853 2.872 8.563 8.585 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.101 8.103 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.418 7.477 calculate_dm_sparse 128 9.5 0.001 0.001 7.046 7.168 rs_scatter_matrices 139 9.7 3.698 4.701 6.755 6.971 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.497 6.542 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=327.116000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2838.636364, yerr=156.657863 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 1.259295E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 38195936. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 86.072 86.073 qs_energies 1 2.0 0.000 0.000 85.583 85.589 ls_scf 1 3.0 0.000 0.000 84.681 84.687 dbcsr_multiply_generic 111 6.7 0.016 0.017 73.381 73.536 multiply_cannon 111 7.7 0.018 0.020 56.555 57.770 multiply_cannon_loop 111 8.7 0.208 0.222 53.115 54.566 ls_scf_main 1 4.0 0.000 0.000 52.717 52.718 density_matrix_trs4 2 5.0 0.002 0.003 47.013 47.104 ls_scf_init_scf 1 4.0 0.000 0.001 28.930 28.931 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.786 27.834 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.660 25.677 mp_waitall_1 11031 10.9 22.774 25.643 22.774 25.643 multiply_cannon_multrec 2664 9.7 8.193 8.866 15.602 17.222 multiply_cannon_sync_h2d 2664 9.7 13.463 15.951 13.463 15.951 make_m2s 222 7.7 0.009 0.011 13.115 13.664 make_images 222 8.7 0.101 0.112 13.093 13.645 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.976 12.678 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.456 9.051 make_images_data 222 9.7 0.005 0.006 7.672 8.188 dbcsr_mm_accdrv_process 4760 10.4 0.509 0.600 7.029 7.994 hybrid_alltoall_any 227 10.6 0.218 1.846 6.587 7.793 dbcsr_mm_accdrv_process_sort 4760 11.4 6.320 7.237 6.320 7.237 calculate_norms 4752 9.8 5.510 6.178 5.510 6.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.076 5.165 mp_sum_l 887 5.1 3.179 4.597 3.179 4.597 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.059 3.691 mp_irecv_dv 6231 10.9 2.043 3.667 2.043 3.667 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.369 3.524 make_images_sizes 222 9.7 0.000 0.000 0.753 3.424 mp_alltoall_i44 222 10.7 0.752 3.424 0.752 3.424 arnoldi_extremal 4 6.8 0.000 0.000 3.349 3.374 arnoldi_normal_ev 4 7.8 0.001 0.002 3.349 3.374 build_subspace 16 8.4 0.009 0.012 3.251 3.254 ls_scf_post 1 4.0 0.000 0.000 3.033 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.858 2.894 dbcsr_special_finalize 555 9.7 0.005 0.006 2.332 2.738 dbcsr_merge_single_wm 555 10.7 0.461 0.586 2.323 2.729 make_images_pack 222 9.7 2.208 2.634 2.210 2.635 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.366 2.593 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.412 2.472 dbcsr_sort_data 658 11.4 2.120 2.456 2.120 2.456 dbcsr_matrix_vector_mult_local 304 10.0 2.050 2.443 2.052 2.445 buffer_matrices_ensure_size 222 8.7 1.762 2.133 1.762 2.133 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.814 1.815 rebuild_ks_matrix 3 7.3 0.000 0.000 1.805 1.806 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 1.805 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.073000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 2.094031E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.062 92.581 92.582 qs_energies 1 2.0 0.000 0.000 92.041 92.059 ls_scf 1 3.0 0.000 0.001 90.713 90.731 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.287 76.672 multiply_cannon 111 7.7 0.029 0.041 53.664 57.860 ls_scf_main 1 4.0 0.000 0.002 56.037 56.042 multiply_cannon_loop 111 8.7 0.116 0.125 50.332 53.644 density_matrix_trs4 2 5.0 0.002 0.005 50.173 50.400 ls_scf_init_scf 1 4.0 0.001 0.003 31.121 31.122 mp_waitall_1 9105 10.9 21.672 30.555 21.672 30.555 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.880 29.975 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.461 27.473 multiply_cannon_multrec 1332 9.7 13.187 16.269 22.550 26.829 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.918 21.138 make_m2s 222 7.7 0.008 0.009 15.764 16.482 make_images 222 8.7 1.589 1.975 15.733 16.453 dbcsr_mm_accdrv_process 4041 10.4 0.285 0.456 8.961 10.596 make_images_data 222 9.7 0.004 0.005 9.124 10.147 dbcsr_mm_accdrv_process_sort 4041 11.4 8.553 10.141 8.553 10.141 hybrid_alltoall_any 227 10.6 0.523 2.485 8.493 9.737 mp_sum_l 887 5.1 5.560 8.721 5.560 8.721 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.249 7.929 mp_irecv_dv 3311 11.0 3.230 7.873 3.230 7.873 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.292 6.977 calculate_norms 2376 9.8 6.004 6.818 6.004 6.818 multiply_cannon_sync_h2d 1332 9.7 4.776 6.152 4.776 6.152 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.095 5.290 arnoldi_extremal 4 6.8 0.000 0.000 4.942 4.956 arnoldi_normal_ev 4 7.8 0.001 0.005 4.942 4.956 build_subspace 16 8.4 0.014 0.021 4.690 4.693 ls_scf_post 1 4.0 0.002 0.016 3.554 3.571 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.264 3.527 ls_scf_store_result 1 5.0 0.000 0.000 3.237 3.365 dbcsr_matrix_vector_mult_local 304 10.0 2.750 3.221 2.752 3.222 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.678 2.764 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.254 2.574 mp_allgather_i34 111 8.7 1.015 2.506 1.015 2.506 make_images_pack 222 9.7 2.026 2.402 2.029 2.404 dbcsr_sort_data 436 11.2 1.829 2.055 1.829 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.923 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.909 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.909 1.912 make_images_sizes 222 9.7 0.000 0.000 0.663 1.908 mp_alltoall_i44 222 10.7 0.663 1.907 0.663 1.907 dbcsr_data_new 4174 10.1 1.619 1.860 1.619 1.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.582000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1769.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 2.715693E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.041 94.460 94.460 qs_energies 1 2.0 0.000 0.000 93.858 93.864 ls_scf 1 3.0 0.000 0.000 92.458 92.467 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.969 77.239 ls_scf_main 1 4.0 0.000 0.000 58.009 58.014 multiply_cannon 111 7.7 0.043 0.103 53.362 57.374 multiply_cannon_loop 111 8.7 0.101 0.119 49.757 53.090 density_matrix_trs4 2 5.0 0.002 0.003 51.973 52.132 mp_waitall_1 7281 11.0 24.085 33.061 24.085 33.061 ls_scf_init_scf 1 4.0 0.000 0.001 30.890 30.893 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.680 29.750 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.287 27.301 multiply_cannon_multrec 888 9.7 12.684 15.170 21.327 24.417 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.389 22.044 make_m2s 222 7.7 0.007 0.008 17.061 18.276 make_images 222 8.7 1.984 2.323 17.022 18.238 make_images_data 222 9.7 0.004 0.005 9.776 10.843 hybrid_alltoall_any 227 10.6 0.623 2.881 9.458 10.683 dbcsr_mm_accdrv_process 3754 10.4 0.244 0.422 8.187 9.445 dbcsr_mm_accdrv_process_sort 3754 11.4 7.812 9.023 7.812 9.023 mp_sum_l 887 5.1 4.991 8.826 4.991 8.826 multiply_cannon_sync_h2d 888 9.7 6.098 7.369 6.098 7.369 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.857 6.895 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.440 6.748 mp_irecv_dv 2335 11.1 2.425 6.694 2.425 6.694 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.550 6.097 arnoldi_extremal 4 6.8 0.000 0.000 5.203 5.222 arnoldi_normal_ev 4 7.8 0.001 0.004 5.203 5.222 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.958 5.139 build_subspace 16 8.4 0.014 0.020 4.902 4.909 calculate_norms 1584 9.8 4.325 4.641 4.325 4.641 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.451 3.792 mp_allgather_i34 111 8.7 1.429 3.766 1.429 3.766 dbcsr_matrix_vector_mult_local 304 10.0 3.010 3.573 3.012 3.575 ls_scf_post 1 4.0 0.000 0.000 3.558 3.566 ls_scf_store_result 1 5.0 0.000 0.000 3.285 3.380 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.887 3.012 make_images_sizes 222 9.7 0.000 0.000 0.944 2.128 mp_alltoall_i44 222 10.7 0.944 2.127 0.944 2.127 make_images_pack 222 9.7 1.809 2.091 1.812 2.094 dbcsr_sort_data 325 11.1 1.881 2.082 1.881 2.082 dbcsr_data_release 9322 10.9 1.299 2.037 1.299 2.037 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.921 1.923 rebuild_ks_matrix 3 7.3 0.000 0.000 1.903 1.905 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.903 1.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.460000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2165.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 3.311141E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.047 97.859 97.860 qs_energies 1 2.0 0.000 0.000 97.226 97.235 ls_scf 1 3.0 0.000 0.000 95.577 95.587 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.697 78.937 ls_scf_main 1 4.0 0.000 0.000 59.285 59.286 multiply_cannon 111 7.7 0.067 0.179 51.732 56.416 density_matrix_trs4 2 5.0 0.002 0.003 52.997 53.130 multiply_cannon_loop 111 8.7 0.115 0.125 46.654 49.346 ls_scf_init_scf 1 4.0 0.001 0.005 33.040 33.042 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.770 31.845 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.309 29.323 mp_waitall_1 6369 11.0 22.356 28.946 22.356 28.946 multiply_cannon_multrec 1332 9.7 14.319 17.073 22.322 25.430 make_m2s 222 7.7 0.008 0.009 20.992 22.463 make_images 222 8.7 3.150 3.615 20.940 22.409 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.830 16.471 make_images_data 222 9.7 0.004 0.005 11.633 13.224 hybrid_alltoall_any 227 10.6 0.803 3.752 10.899 12.681 dbcsr_mm_accdrv_process 3641 10.4 0.214 0.408 7.641 9.192 dbcsr_mm_accdrv_process_sort 3641 11.4 7.266 8.768 7.266 8.768 mp_sum_l 887 5.1 4.383 8.409 4.383 8.409 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.390 6.685 multiply_cannon_sync_h2d 1332 9.7 5.524 6.208 5.524 6.208 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.111 5.998 mp_irecv_dv 3229 10.9 2.089 5.913 2.089 5.913 arnoldi_extremal 4 6.8 0.000 0.000 5.640 5.662 arnoldi_normal_ev 4 7.8 0.001 0.004 5.640 5.662 build_subspace 16 8.4 0.014 0.021 5.279 5.286 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.725 5.272 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.597 4.753 mp_allgather_i34 111 8.7 2.171 4.638 2.171 4.638 calculate_norms 2376 9.8 4.193 4.527 4.193 4.527 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.762 4.024 dbcsr_matrix_vector_mult_local 304 10.0 3.182 3.651 3.184 3.653 dbcsr_sort_data 658 11.4 3.048 3.479 3.048 3.479 ls_scf_post 1 4.0 0.000 0.000 3.252 3.259 dbcsr_special_finalize 555 9.7 0.006 0.007 2.788 3.234 dbcsr_merge_single_wm 555 10.7 0.532 0.661 2.779 3.225 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.120 3.168 ls_scf_store_result 1 5.0 0.000 0.000 2.973 3.036 dbcsr_data_release 10477 10.7 1.597 2.430 1.597 2.430 dbcsr_finalize 304 7.8 0.050 0.061 1.817 2.005 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.986 1.987 rebuild_ks_matrix 3 7.3 0.000 0.000 1.963 1.965 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.963 1.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.860000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 4.680999E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.074 99.824 99.825 qs_energies 1 2.0 0.000 0.000 98.999 99.005 ls_scf 1 3.0 0.000 0.000 97.070 97.080 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.380 77.634 ls_scf_main 1 4.0 0.000 0.000 62.291 62.292 multiply_cannon 111 7.7 0.081 0.133 54.714 59.983 density_matrix_trs4 2 5.0 0.002 0.003 54.914 55.002 multiply_cannon_loop 111 8.7 0.069 0.076 50.112 51.862 mp_waitall_1 5436 11.0 25.543 31.291 25.543 31.291 ls_scf_init_scf 1 4.0 0.001 0.001 31.203 31.208 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.844 29.884 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.646 27.658 multiply_cannon_multrec 444 9.7 14.125 16.465 21.286 23.553 make_m2s 222 7.7 0.006 0.007 17.889 20.548 make_images 222 8.7 3.727 4.429 17.826 20.485 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.538 14.965 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.860 13.744 make_images_data 222 9.7 0.003 0.004 9.936 12.707 hybrid_alltoall_any 227 10.6 0.791 3.776 9.632 12.300 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.346 6.852 7.990 multiply_cannon_sync_h2d 444 9.7 6.532 7.989 6.532 7.989 dbcsr_mm_accdrv_process_sort 3003 11.4 6.534 7.640 6.534 7.640 mp_allgather_i34 111 8.7 2.811 7.129 2.811 7.129 arnoldi_extremal 4 6.8 0.000 0.000 6.590 6.604 arnoldi_normal_ev 4 7.8 0.001 0.004 6.590 6.604 build_subspace 16 8.4 0.015 0.019 6.185 6.197 mp_sum_l 887 5.1 2.839 5.169 2.839 5.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.653 4.791 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.587 4.680 dbcsr_matrix_vector_mult_local 304 10.0 3.620 4.044 3.622 4.046 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.689 3.978 mp_irecv_dv 1241 11.2 1.671 3.950 1.671 3.950 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.917 3.856 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.702 3.810 calculate_norms 792 9.8 3.558 3.691 3.558 3.691 make_images_sizes 222 9.7 0.000 0.000 1.281 3.590 mp_alltoall_i44 222 10.7 1.281 3.590 1.281 3.590 ls_scf_post 1 4.0 0.000 0.000 3.576 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.345 3.404 dbcsr_finalize 304 7.8 0.062 0.078 2.202 2.270 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.163 2.164 rebuild_ks_matrix 3 7.3 0.000 0.000 2.131 2.132 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.131 2.132 dbcsr_merge_all 275 8.9 0.475 0.527 2.050 2.107 dbcsr_data_release 10123 10.8 1.355 2.010 1.355 2.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.825000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3678.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_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 8.713413E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.104 0.151 110.710 110.711 qs_energies 1 2.0 0.000 0.000 109.140 109.152 ls_scf 1 3.0 0.000 0.000 106.175 106.187 dbcsr_multiply_generic 111 6.7 0.023 0.028 79.082 79.201 ls_scf_main 1 4.0 0.000 0.000 66.948 66.949 density_matrix_trs4 2 5.0 0.002 0.003 57.441 57.498 multiply_cannon 111 7.7 0.174 0.244 51.003 52.956 multiply_cannon_loop 111 8.7 0.068 0.070 47.467 48.470 ls_scf_init_scf 1 4.0 0.001 0.001 35.486 35.486 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.793 33.813 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.280 30.292 mp_waitall_1 4527 11.1 23.215 27.607 23.215 27.607 make_m2s 222 7.7 0.007 0.007 24.386 25.414 make_images 222 8.7 4.603 5.015 24.276 25.302 multiply_cannon_multrec 444 9.7 17.919 18.562 22.864 23.406 hybrid_alltoall_any 227 10.6 1.667 3.631 13.244 16.096 make_images_data 222 9.7 0.003 0.004 13.478 16.093 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.026 11.863 multiply_cannon_sync_h2d 444 9.7 8.846 8.892 8.846 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.454 7.464 arnoldi_normal_ev 4 7.8 0.002 0.008 7.454 7.464 build_subspace 16 8.4 0.026 0.036 6.888 6.898 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.736 6.004 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.373 5.512 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.395 5.488 dbcsr_mm_accdrv_process 1814 10.4 0.272 0.974 4.774 5.323 dbcsr_matrix_vector_mult_local 304 10.0 4.878 5.175 4.880 5.177 dbcsr_mm_accdrv_process_sort 1814 11.4 4.210 4.350 4.210 4.350 make_images_sizes 222 9.7 0.000 0.000 1.544 3.858 mp_alltoall_i44 222 10.7 1.544 3.858 1.544 3.858 ls_scf_post 1 4.0 0.000 0.000 3.741 3.752 mp_allgather_i34 111 8.7 1.155 3.640 1.155 3.640 ls_scf_store_result 1 5.0 0.000 0.000 3.430 3.437 calculate_norms 792 9.8 3.241 3.279 3.241 3.279 dbcsr_finalize 304 7.8 0.082 0.090 3.105 3.192 dbcsr_merge_all 275 8.9 0.891 0.914 2.889 2.970 dbcsr_complete_redistribute 5 7.6 1.430 1.470 2.854 2.953 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.935 2.935 dbcsr_data_release 12724 10.6 2.335 2.892 2.335 2.892 matrix_ls_to_qs 2 6.0 0.000 0.000 2.499 2.613 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.589 2.591 rebuild_ks_matrix 3 7.3 0.000 0.000 2.522 2.524 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.005 2.522 2.524 dbcsr_sort_data 325 11.1 2.450 2.522 2.450 2.522 dbcsr_new_transposed 4 7.5 0.243 0.252 2.395 2.415 mp_alltoall_d11v 48 9.2 2.261 2.309 2.261 2.309 dbcsr_frobenius_norm 74 6.6 2.059 2.134 2.198 2.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.711000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6850.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/186428f4e848f883d21da2d2b78d036968ebedff_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.143488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 812824902. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.114 339.483 339.484 qs_mol_dyn_low 1 2.0 0.006 0.080 338.215 338.241 qs_forces 5 3.8 0.006 0.040 337.990 337.993 qs_energies 5 4.8 0.005 0.053 334.264 334.289 scf_env_do_scf 5 5.8 0.000 0.002 315.698 315.701 scf_env_do_scf_inner_loop 105 6.6 0.004 0.038 272.325 272.355 qs_scf_new_mos 105 7.6 0.000 0.001 214.701 215.083 qs_scf_loop_do_ot 105 8.6 0.001 0.001 214.701 215.083 ot_scf_mini 105 9.6 0.003 0.005 204.339 204.662 dbcsr_multiply_generic 1445 12.2 0.124 0.131 142.691 144.447 velocity_verlet 4 3.0 0.002 0.028 130.816 130.818 multiply_cannon 1445 13.2 0.280 0.297 119.113 123.292 multiply_cannon_loop 1445 14.2 2.400 2.521 115.868 120.230 qs_ot_get_p 112 10.4 0.001 0.002 107.962 108.274 qs_ot_p2m_diag 40 11.0 0.020 0.031 94.713 94.814 cp_dbcsr_syevd 40 12.0 0.003 0.039 91.078 91.082 cp_fm_syevd 40 13.0 0.000 0.004 84.219 84.393 cp_fm_redistribute_end 40 14.0 39.033 77.871 39.045 77.880 cp_fm_syevd_base 40 14.0 38.803 77.649 38.803 77.649 ot_mini 105 10.6 0.001 0.003 64.840 64.969 mp_waitall_1 488190 16.1 42.992 54.945 42.992 54.945 multiply_cannon_multrec 69360 15.2 30.773 36.893 38.691 44.406 init_scf_loop 7 6.6 0.000 0.007 43.305 43.309 qs_ot_get_derivative 55 11.6 0.001 0.001 42.868 43.052 multiply_cannon_metrocomm3 69360 15.2 0.194 0.203 27.932 42.875 prepare_preconditioner 7 7.6 0.000 0.000 37.102 37.139 make_preconditioner 7 8.6 0.000 0.003 37.102 37.139 rebuild_ks_matrix 110 8.4 0.000 0.000 36.172 36.379 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.024 36.171 36.379 multiply_cannon_sync_h2d 69360 15.2 29.465 34.044 29.465 34.044 qs_ks_update_qs_env 112 7.6 0.001 0.001 33.249 33.442 qs_rho_update_rho_low 110 7.6 0.001 0.005 29.253 29.546 calculate_rho_elec 110 8.6 0.030 0.034 29.252 29.544 make_full_inverse_cholesky 7 9.6 0.000 0.006 27.558 27.626 rs_pw_transfer 690 11.5 0.011 0.020 25.808 27.122 density_rs2pw 110 9.6 0.006 0.010 22.182 23.528 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.227 23.471 apply_single 62 13.6 0.000 0.000 23.227 23.471 ot_new_cg_direction 55 11.6 0.001 0.003 21.205 21.207 cp_fm_cholesky_invert 7 10.6 18.534 18.556 18.534 18.556 qs_ot_get_orbitals 105 10.6 0.001 0.001 16.016 16.379 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.366 15.506 init_scf_run 5 5.8 0.000 0.001 15.434 15.436 scf_env_initial_rho_setup 5 6.8 0.000 0.005 15.434 15.435 pw_transfer 1645 12.4 0.084 0.110 14.513 14.909 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 14.369 14.772 mp_sum_l 4764 12.2 13.391 14.577 13.391 14.577 sum_up_and_integrate 60 10.3 0.028 0.030 14.416 14.434 integrate_v_rspace 60 11.3 0.002 0.002 14.388 14.407 multiply_cannon_metrocomm1 69360 15.2 0.096 0.104 5.624 13.102 calculate_dm_sparse 110 9.5 0.000 0.001 12.041 12.295 check_diag 80 13.5 8.588 8.861 12.030 12.188 fft_wrap_pw1pw2_240 915 15.0 1.193 1.291 11.720 12.098 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.672 11.757 qs_vxc_create 110 10.4 0.004 0.046 11.418 11.481 rs_pw_transfer_RS2PW_30 110 11.6 1.584 1.660 10.670 11.480 mp_sendrecv_dv 168740 12.6 11.028 11.176 11.028 11.176 make_m2s 2890 13.2 0.083 0.090 9.757 10.452 make_images 2890 14.2 0.244 0.265 9.647 10.343 fft3d_pb 915 16.0 2.381 2.692 9.769 10.255 potential_pw2rs 60 12.3 0.003 0.003 9.830 9.903 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 9.790 9.808 acc_transpose_blocks 69360 15.2 0.359 0.384 8.850 9.394 calculate_first_density_matrix 1 7.0 0.000 0.006 9.309 9.331 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.142 9.218 make_full_single_inverse 7 9.6 0.001 0.005 9.039 9.076 dbcsr_mm_accdrv_process 154766 15.8 4.137 4.297 7.793 8.624 cp_fm_cholesky_decompose 14 10.2 8.383 8.396 8.383 8.396 xc_rho_set_and_dset_create 110 12.4 0.076 0.096 7.856 8.113 xc_vxc_pw_create 60 11.3 0.039 0.049 7.738 7.801 mp_alltoall_z22v 2340 17.7 7.027 7.605 7.027 7.605 acc_transpose_blocks_kernels 69360 16.2 0.882 0.924 6.913 7.374 xc_pw_derive 510 13.4 0.006 0.007 7.038 7.114 mp_alltoall_d11v 1300 13.8 6.121 6.976 6.121 6.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=339.484000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 186428f4e848f883d21da2d2b78d036968ebedff Summary: empty Status: OK