[Ffmpeg-devel] MMX/MMX2 and SSE optimizations for H.264 decoding

Loren Merritt lorenm
Mon Sep 19 17:27:34 CEST 2005

On Mon, 19 Sep 2005, Martin Boehme wrote:

> OK, so it sounds as if there's far less optimization in FFmpeg's H.264 on PPC 
> than on x86... which probably isn't a surprise...

We have mmx for 16x16..4x2 motion compensation, dct, and deblocking.
We have altivec only for luma 16x* motion compensation, and whatever
misc functions are reused from other codecs.

--Loren Merritt

More information about the ffmpeg-devel mailing list