Using the ARM Cortex alone, or with the help of the DSP + associated coprocessors? The Big Buck Bunny demo everybody always sees is decoded with the help of a lot of NEON optimizations by omapfb. It does not have sound and runs entirely on the Cortex. IVA decoding isn't present yet, but will be coming in the future, and it will also be able to handle 720p. #beagle on Freenode can probably offer some clearer details.
Using the ARM Cortex alone, or with the help of the DSP + associated coprocessors?