<!DOCTYPE html>
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body>
<p><font face="serif">Class -- I have gotten several pleas for help
today (I know, you have a project due tomorrow). Due to several
reasons I cannot get to campus today. I know some of you need a
really good grade on this project and, ror that reason, I am
going to extend the deadline for this project to next MONDAY and
we will have tomorrow in class and Friday in class to finish up
any nasty details. I will also be in Godsey tomorrow morning
working in either the research lab (GSC 405) or the
classroom/server room (GSC 218). </font></p>
<p><font face="serif">One thing that I do want to mention -- don't
stress about using PAPI for the determining the speedup for the
iterative solver - just assume that it is using the same number
of iterations as the direct solver. Experience in the last few
weeks has shown me that some of the double precision results
from PaPi can't be trusted and that several additional checks
have to be included in your PaPi calls to verify that you are
actually getting the flops counted properly. <br>
</font></p>
<p><font face="serif">I<br>
</font></p>
<div class="moz-signature">-- <br>
<b><i>Andrew J. Pounds, Ph.D.</i></b><br>
<i>Professor of Chemistry and Computer Science</i><br>
<i>Director of the Computational Science Program</i><br>
<i>Mercer University, Macon, GA 31207 (478) 301-5627</i></div>
</body>
</html>