[petsc-dev] DD FAQ

[Barry Smith]

On Apr 26, 2010, at 9:03 AM, Jed Brown wrote:

> The FAQ mentions PCNN, but the advice "-pc_type nn" is completely
> inadequate to actually use this method, and the referenced example  
> does
> not exist.  I have removed it's mention from the "Domain  
> Decomposition"
> answer, I think it needs a working example in order to be mentioned
> there.  Also, ex2 -pc_type nn -mat_type fails with a seg-fault, I
> haven't investigated why.
>
> Barry, PCEXOTIC does have a working example, should it be mentioned
> here, or do you consider it to be a toy that users needn't bother  
> trying
> to use.

    Done

>
> As for future DD directions, have you thought about what is  
> necessary to
> provide quasi-generic support for optimized Schwarz.  I was only  
> vaguely
> aware of this method until recently, but this approach seems to offer
> significant potential, especially for some indefinite problems.

    I assume you mean where you use "optimized" linear combinations of  
solutions in overlapped regions? I am only vaguely aware as well. I've  
attempted to ignore them all these years. I don't know how to tune  
them. We currently use VecScatter to put in the overlapped values; now  
we'd need to multiple each value by some scale factor before putting  
in. I hate to increase the complexity of VecScatter even more by  
adding this support but we could.

    Barry

>
> Jed