In this paper we focus on algorithms and reconfigurable multi-core architectures for streaming digital signal processing (DSP) applications. The multi-core concept has a number of advantages: (1) depending on the requirements more or fewer cores can be switched on/off, (2) the multi-core structure fits well to future process technologies, more cores will be available in advanced process technologies, but the complexity per core does not increase, (3) the multi-core concept is fault tolerant, faulty cores can be discarded and (4) multiple cores can be configured fast in parallel. Because in our approach processing and memory are combined in the cores, tasks can be executed efficiently on cores (locality of reference). There are a number of application domains that can be considered as streaming DSP applications: for example wireless baseband processing (for HiperLAN/2, WiMax, DAB, DRM, and DVB), multimedia processing (e.g. MPEG, MP3 coding/decoding), medical image processing, colour image processing, sensor processing (e.g. remote surveillance cameras) and phased array radar systems. In this paper the key characteristics of streaming DSP applications are highlighted, and the characteristics of the processing architectures to efficiently support these types of applications are addressed. We present the initial results of the Annabelle chip that we designed with our approach.
|Workshop||Tenth International Workshop on System-Level Interconnect Prediction, SLIP 2008|
|Period||5/04/08 → 6/04/08|
|Other||5-6 April 2008|
- EC Grant Agreement nr.: FP6/001908
- CAES-EEA: Efficient Embedded Architectures