On Mon, Oct 31, 2011 at 8:15 AM, Gong Ding <span dir="ltr"><<a href="mailto:gdiso@ustc.edu">gdiso@ustc.edu</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Yes, modern direct solver do block factorization as super node method.<br>
Can ASM type preconditioner use superlu ILUT on each subdomain?<br></blockquote><div><br></div><div>Yes.</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
At preset, my solver (BCGS) only works with ASM+ILU(1) on a "smart partitioned" mesh.<br>
<div><div></div><div class="h5"><br>
<br>
> The ILUTP in SuperLU (4.2 and up) exploits blocking (supernodes) in<br>
><br>
> the (approximate) L & U factors. Note the block boundary is<br>
><br>
> discovered on the fly, which is usually larger than the block size<br>
><br>
> from the input matrix A. So the efficiency is pretty good. It uses<br>
><br>
> threshold dropping with partial pivoting, numerically is quite good<br>
><br>
> too. It's available only in the serial version.<br>
><br>
><br>
><br>
> Sherry Li<br>
><br>
><br>
><br>
><br>
><br>
> On Sun, Oct 30, 2011 at 11:18 PM, Jed Brown <<a href="mailto:jedbrown@mcs.anl.gov">jedbrown@mcs.anl.gov</a>> wrote:<br>
><br>
> > On Sun, Oct 30, 2011 at 23:50, Gong Ding <<a href="mailto:gdiso@ustc.edu">gdiso@ustc.edu</a>> wrote:<br>
><br>
> >><br>
><br>
> >> I had also considered use max block size. But it is not efficient here.<br>
><br>
> >> Hope petsc support nonuniform block size one day.<br>
><br>
> ><br>
><br>
> > Inodes do a sort of partial blocking. There has to a clear performance<br>
><br>
> > benefit to explicit variable blocking in order to justify the implementation<br>
><br>
> > and interface complexity. I have not yet seen a demonstration of this.<br>
><br>
><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener<br>