PALMING OFF system-on-chip (SoC) audio functions to secondary chips will not only improve audio performance but also battery life, claims Wolfson.
Wolfson's chief commerial officier, Andy Brannan made the comments while demonstrating the firm's new WM0010 digital signal processing unit (DSP) that aims to only wake the main SoC to send a burst of data. The WM0010 offers a 100db signal to noise ratio, and Brannan claimed Wolfson will be making considerable strides in 2011 to increase that into the realm of true hi-fi quality.
According to Brannan, the use of DSPs such as Wolfson's will save battery power as larger, more power hungry SoCs draw significantly more power than relatively simple DSP units.
In a demonstration for The INQUIRER, the WM0010 was shown to be playing an audio file that was part buffered, with the main SoC being woken up only every seven seconds. The playback was faultness, with no difference between when the SoC was pushing the data to the audio chip and when it was being played from the buffer.
The aim, Brannan said, is to "not wake up the SoC" very often and while SoC chips have increased in power efficiency, they still have significantly higher power draw compared to DSPs.
Brannan said that manufacturers can not only make use of multi-chip audio designs to extend battery life but can significantly improve audio quality.
The notion of splitting parts of the SoC off represents a reversal of recent design strategy. It has been all about aggregating functions, however as certain chips remain more power efficient than larger SoCs it makes sense to use them where possible.
Brannan said that Wolfson's WM0010 DSP chip is already being sampled and products using it are expected soon. µ