[Ffmpeg-devel] Fixed vs. Floating Point AAC
Thu Mar 9 11:26:51 CET 2006
On Thu, 2006-03-09 at 01:27 +0000, M?ns Rullg?rd wrote:
> > for the athlon the timings arent clear from the docs i have, only that
> > 32*32->32 seems 1/4 and 32*32->64 worse then 1/6 if the high value is used
> > and FMUL >=1/4, also note fmul is direct path imul vector path so imul
> > cannot excute with anything else together while fmul can
> How about doing a 64-bit version for true 64-bit CPUs? Low-end
> embedded systems often run 64-bit MIPS without floating-point.
How does athlon64 (and EM64T) perform in this area? Any data? I'd expect
it can do a decent 32x32 multiplication...
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 2771 bytes
Desc: not available
More information about the ffmpeg-devel