<div class="gmail_quote">On Fri, Dec 3, 2010 at 22:15, Hong Zhang <span dir="ltr"><<a href="mailto:hzhang@mcs.anl.gov">hzhang@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div id=":t2">Yes, old ICC was buggy. Unlike Cholesky which reveals error immediately,<br>
incorrect ICC may still gives convergence. I roughly recall that<br>
comparing convergence of<br>
icc for aij and sbaij, I fixed bug in one of the routines (likely for sbaij).<br>
What block size of sbaij do you use?<br></div></blockquote><div><br></div><div>This was for bs=2. Note that I plain AIJ was requiring a shift and SBAIJ(2) had negative pivots (but didn't do anything about them). So they would have had to both be buggy.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div id=":t2">
While changing data structures in matrix factorization routine, I<br>
cleaned or rewrote<br>
part of factorization routines, mainly for bs=1. bs>1 might need more work, no<br>
guarantee there :-(</div></blockquote></div><br><div>The new one is working, I have identical results with AIJ and SBAIJ(2). Neither of them need a shift. This is good, except that I no longer have a nice explanation for why ICC performs so poorly on this problem when used with 1-level ASM (no coarse level). The difference between direct subdomain solves and incomplete solves, even with small subdomains, is an order of magnitude for ASM, but negligible when there is a coarse level. I know this is not overly weird, but having ICC produce pivots that needed shifting (despite the matrix being SPD) was a tidy explanation that apparently I have to discard in favor of "incomplete factorization is unpredictable". :-(</div>
<div><br></div><div>Thanks for running through the history with me.</div><div><br></div><div>Jed</div>