Benchmarking against AVTP#


  • complete case description: outlet Dirichlet 1300K, everything else 300K, fixed timestep

  • show how to run FEniCS in kraken (make a how to)

  • speak about conda problems + fenics problems when too little dof per cpu


  • results are outdated: new fenics version is much faster

  • design better the benchmarks

For a coarse mesh (~150k cells):


For a fine mesh (~10M cells):