[Ffmpeg-devel] Re: about mmx instructions

thomas.kunlin at free.fr thomas.kunlin
Thu Sep 1 17:15:30 CEST 2005


Thank you very much for your reply !

> anyone got a nicer derivation/proof?
> or even a faster implementation?

A x2 speedup could be achieved by using the SSE2 integer
instructions (128bit registers instead of 64bit for MMX),
using the very same algorithm. I am working on it at this
moment.

Thomas





More information about the ffmpeg-devel mailing list