[ODE] More speed???

Adam D. Moss adam at gimp.org
Thu Nov 6 16:21:31 MST 2003


DjArcas wrote:
 >>>>           -These two function can be SIMDize
 >>>>            easily. I have not tried it yet but I'd bet it will make
 >>>>            original Step() much faster.
 >>
 >>MCM> Definitely!  Russ has suggested this a couple times.
 >>                I'm a half way doing this. :)
 >
 >
 > don't forget us AMD owners ;)

SIMD is a general term, but you're right, I imagine it'll first
be SSE like dSetZero().  :)   ODE's source is available, so
anyone who knows 3dnow/etc can always add support separately...
but runtime autodetection glue so often gets forgotten...

I wish there was a good cross-platform solution for SIMD (does
anyone know any?).  GCC 3.4 (&3.3?) seems to have some low-level
cross-platform vector builtins, but I don't know how practical
they work out to be.  I don't think they begin to cover runtime
arch detection and code-switching though.

--Adam
-- 
Adam D. Moss   . ,,^^   adam at gimp.org   http://www.foxbox.org/   co:3



More information about the ODE mailing list