[petsc-dev] Non-scalable matrix operations

[Mark F. Adams]

On Dec 23, 2011, at 2:39 PM, Jed Brown wrote:

> On Fri, Dec 23, 2011 at 13:37, Mark F. Adams <mark.adams at columbia.edu> wrote:
> Also, my algorithm exploits static partitions where you know who is processing your ghost nodes (so you can avoid stepping on each other).  This might not be available.  Coloring looks OK for 7 point stencils but the number of colors gets very large at higher order.  You need about 13 colors for a 3D hex mesh (not very high order at all).  But inside the compute node there are sets of vertices that need processing and you can do whatever you want on these sets, so if you need to color then so be it.
> 
> I think the number of colors is pretty much hopeless if you use a stable Stokes element like Q2-P1disc.

Yes, thats why I think you want to try to bring this algorithm down to the "thread" level.  You basically need subdomains that are say a few, or at very least one, stencil width across.  G-S is complicated but it does use only one vector instead of two for additive methods and it can have good math properties.  So I'm not sure if its worth pursuing but its something to keep in mind.

Mark  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20111223/9ad2f4f6/attachment.html>