On Sat, 2006-04-01 at 00:19 +0100, Brian Gough wrote: > > Until LAPACK has transpose arguments to make a 'LAPACK trick' possible > it's a lost cause and is probably easier for people call LAPACK > directly on a case-by-case basis.
I am hoping that the new work will address this problem. Check the message I sent earlier with the link to the 2004 Dongarra et al. proposal. I'm sure you've seen this; I think we talked about it. We need to figure out what they are doing, and act accordingly. They may be solving all these problems. -- Gerard Jungman <[EMAIL PROTECTED]> Los Alamos National Laboratory
