In recent years, the market for DSP-capable processors has grown tremendously. In response to the demand for faster, cheaper, smaller, more energy-efficient DSP processors, DSP processor vendors have developed processor architectures that depart significantly from those of “traditional” DSPs. In addition, microcontroller and CPU vendors have begun to add DSP enhancements to what were originally traditional general-purpose architectures, creating hybrid processors that may challenge the performance of dedicated DSPs.

The large selection of processors equipped with DSP capabilities is clearly an advantage for system designers, since there are now processors suitable for a wide variety of applications. However, the somewhat bewildering array makes it both difficult and time-consuming to select the best processor for an application. In this paper, we will discuss the methodology we have developed for evaluating processor performance in DSP applications, a main criterion for processor selection. We will summarize our latest findings using this methodology, and provide some insight into how and why the newest processors perform as they do on our benchmarks.