[tremor] Ogg/Vorbis report, FFT optimizations

[Segher Boessenkool]

> Secondly, as far as I understand, when you compare the performance of 
> the old MDCT with your new FFT based implementation, you are 
> comparing:
> A pure C implementation of a MDCT written specifically for processors 
> with a 32*32bit multiplication instruction, but complied for a 
> processor with only a 16*16bit multiplication instruction.
> With:
> A FFT based MDCT implementation where the FFT is highly asm optimised 
> by Texas Instruments for this particular DSP.
> Not a very fair comparison.

That depends on your problem domain.  From his point of view, it 
definitely
_is_ a fair comparison.

> The obvious fact that the original Tremor codebase is neither suited 
> nor intended for 16bit processors shows clearly in table 7.4. 
> According to that table, Tremor require 107MIPS to decode at 112kbps 
> on your DSP. Tremor runs fine on a ~60MIPS ARM7.

...Which is pretty pathetic performance as well, really.

> In your paper, you suggest a system consisting of a ARM processor for 
> parsing and pre-decoding, and a DSP (presumably the 16bit TI DSP you 
> use) for decoding. This sounds rather silly to me, since Tremor is 
> specifically designed to run on 32bit ARM CPUs.

Distributing the workload over several cores, each designed to do the
kind of task they are doing, makes for quite nice power savings, if
nothing else.  Very important for mobile things.

> I also spotted a small technical error on page 35.. It says the DSP 
> requires all 16bit words to be aligned on 2-byte boundarys, all 32bit 
> words on 4-byte boundarys, and that this is not the case on ARM 
> processors..
> Words must allways be properly aligned on all processors. This is not 
> specific for your DSP.

Words do *not* have to be aligned on al processors.

> It would be very interesting to see the results of your FFT based 
> approach on a ARM7 machine!

Yes.  Or on mainstream desktop machines, too.

<p>Looking forward to read the report (I downloaded it, just don't have 
time
to read it right now),

<p>Segher

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'tremor-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.