What do you guys thing about the P4
lindahl at conservativecomputer.com
Fri Apr 13 10:39:14 PDT 2001
On Fri, Apr 13, 2001 at 10:24:48AM -0700, Phillip D. Matz wrote:
> Which brings up another question. My current impression was that SSE/SSE2
> is only single-precision while G98 (and probably most computational
> chemistry programs) requires double-precision -> if this is correct then SSE
> doesn't help, rather, it invalidates ones G98 results. Would anyone care to
> set me straight?
Starting with the P4, there is double-precision in SSE. That's part of
why it's suddenly important to get a compiler that really understands
the P4 -- and Intel's compiler is the only one at the moment.
More information about the Beowulf