Hi Tariq, > is there a reason example 32 > (http://www.dealii.org/developer/doxygen/deal.II/step_32.html) was not > included in the example list accessible via the main deal-page ?
That is most likely, because step-32 is not done yet. We are still working on fixing some bugs in 3d, improving the stabilization (especially in 3d), and the solver performance... > I'd really like to use the beautiful combined threading/mpi approach shown > there, but I wonder if there are any hidden problems. Well, the combination of threading and MPI (often called hybrid mode) is not as trivial as it looks. We haven't made many tests with that yet. Hybrid mode only makes sense when a large percentage of the algorithms are parallelized to take advantage of multithreading and MPI. This is only partly the case with the Trilinos solvers right now for example (at least some parts seem to use OpenMP but can also not detect the right number of cores to use). So, it is something we want to look into, but I fear, that the performance of the pure MPI variant is preferable right now. But if you want to experiment, you are welcome to check out the development version and try it out! Best, Timo -- Timo Heister http://num.math.uni-goettingen.de/~heister _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
