p1 Next-Generation Consumer Audio Application Specific Embedded Processor

Abstract—For next-generation audio applications, the dominant
trends are much higher sample rate, larger word length
and more audio channels for playback audio data.

I. INTRODUCTION
Most of current audio solutions are based on DSPs or
embedded processors to support various audio applications
flexibly.

However, for next-generation audio applications with
higher sampling resolutions, such as 24-bit or even 32-bit,
current processor architectures are less efficient. One example
is that most DSPs or embedded processors only support 48-
bit or 64-bit accumulators and 16×16 or 16×32 multiplications.

(page 1 col 2)
Another consideration for next-generation audio application
is the coding method of audio data

The on-chip memory requirement is another factor that
mainly affects the performance and cost of audio solutions

(page 2)
The approach to co-design an application-specific processor
to meet the computational needs of a set of applications
of interest has been applied quite successfully in recent
years

To evaluate the performance of the proposed audio processor,
typical audio processing algorithms for next-generation
audio applications have been used as audio specific benchmarks.

The rest of the paper is organized as follows.

II. AUDIO KERNEL BENCHMARK

For audio data compression and encoding, various audio
standards have been applied in consumer audio equipments.

(p2 col 2)
For multichannel audio applications, the audio channel
downmixing technology is usually used to produce surround
sound effect with stereophonic sound playback equipments,
such as high fidelity headphones.