On Mon, Dec 30, 2013 at 11:27 PM, Victor Eijkhout
<[email protected]>wrote:

>
>  On Dec 30, 2013, at 10:49 PM, Matthew Knepley <[email protected]>
>  wrote:
>
> How hard is it to make your code do 2D DAs?
>
>
>  Probably not hard. But
> 1. This is not intended as a production code.
> 2. I am mostly concerned with the communication part.
> 3. Ergo, the local computation part of each task is hard-coded.
> (Determination of the halo and all that.) It will take some serious
> compiler technology to automate that and right now I don't even pretend to
> make an attempt at that.
>

I don't care at all about the local computation. The communication setup
for DA is one of the uglist
parts of PETSc. Automating that would be a serious advance.

   Matt


>
>  Victor.
> --
> Victor Eijkhout, 512 471 5809 (w)
> Texas Advanced Computing Center,
> The University of Texas at Austin
>
>
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener

Reply via email to