On Mon, Jul 11, 2005, Robert Duke wrote:
> Okay, I looked at the apple G5 pmemd performance optimizations...
> There are significant changes
> everywhere in the 9 codebase, and adapting this stuff to 9 will probably
be
> nontrivial. I would recommend we let apple support this level of patch
> themselves, releasing it on their applications website or whatever;
Sounds fine to me. We did recently post a "efficiency patch" for Itanium2
that was processor-specific. But in that case, we have Roberto to provide
support for it. (And here as well, the transition to Amber 9 will be
non-trivial....)
Whether we post this on our site, or just have a link to Apple's site, we
should definitely indicate that this is unsupported...if people have
problems,
they will probably just be directed to revert to the "real" pmemd code.
On the other hand, there are some very experienced Amber people out there
who
have access to G5 clusters, and if we can make their simulations faster,
that
would be a Good Thing.
...dac
Received on Wed Apr 05 2006 - 23:49:54 PDT
