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

[Martin Boehme]

Romain Dolbeau wrote:
> Loren Merritt wrote:
> 
>> we have:
>> put_h264_qpel_16_mc*_altivec
>> put_h264_chroma_mc8_altivec
>>
>> we do not have equivalents of:
>> put_h264_qpel_8_mc*_mmx2
>> put_h264_qpel_4_mc*_mmx2
>> h264_*_loop_filter_*_mmx2
>> ff_h264_idct_add_mmx2
> 
> 
> The additional question is, are we going to be bitten by
> Amdahl's Law ? i.e. are these important enought during
> the decoding process to justify optimizing them ?

The performance difference between H.264 on PPC and x86 seems to 
indicate that they are... but of course they should be profiled...

Martin

-- 
Martin B?hme
Inst. f. Neuro- and Bioinformatics
Ratzeburger Allee 160, D-23538 Luebeck
Phone: +49 451 500 5514
Fax:   +49 451 500 5502
boehme at inb.uni-luebeck.de