Arm Releases Highly Optimized Media Code

[daemonaholic]

Just thought someone might want to take a look at this for any media player development :

"ARM Releases AAC, MP3, MPEG-4, H.264 and FFT OpenMAX DL Libraries, Highly Optimized for Cortex-A8/NEON and ARM11 Processors"
http://www.arm.com/news/21181.html

 

[Exophase]

Nice but using the DSP is the best way to go, most likely. Although this might be more of an accessible option.

 

[icurafu]

That's awesome.

It's over a month old, but I never heard about it.

QUOTE
"and the license granted to these functions entitles the user to integrate them into their own products at no cost."

 

[Chip]

Nice but using the DSP is the best way to go, most likely. Although this might be more of an accessible option.

Exactly. Codecs optimized for the DSP would be ideal, but this ARM code will likely give us full speed HD h.264 playback right off the bat. Depending on how efficient it is, it might not even be worth the trouble of making DSP-based codecs.

 

[Sphinxter]

Must be something to be done on the dsp, shame just going to waste hanging there. Might the dsp lower power consumption still make it worthwhile?

 

[Tinnus]

Depending on how efficient it is, it might not even be worth the trouble of making DSP-based codecs.

Battery life?

 

[.Gogeta§§J4BR.]

nice to see we wont have to rely on free-time programmers to wait for a good media player to come out (ie: tcpmp for palm os)

 

[Chip]

Depending on how efficient it is, it might not even be worth the trouble of making DSP-based codecs.

Battery life?

Hence the "Depending on how efficient." If the power savings of DSP-native codecs over ARM-native codecs is only 10-15%, then it might not be worth the effort of re-writing them specifically for the DSP. I have no idea how much processing power the ARM code uses, how much less power DSP code might use, or just how much of a chore re-writing the codecs would be. That's why I was sure to qualify the statement

 

[daemonaholic]

Nice but using the DSP is the best way to go, most likely. Although this might be more of an accessible option.

Exactly. Codecs optimized for the DSP would be ideal, but this ARM code will likely give us full speed HD h.264 playback right off the bat. Depending on how efficient it is, it might not even be worth the trouble of making DSP-based codecs.


Exactly what I was thinking when I ran across the article. Sure, the ideal goal is to fully utilize the OMAP3 platform in the most efficient manner...but in the meantime, we can use some of the power from the Cortex-A8 core to handle multimedia .