Code Transformations for Embedded Reconfigurable Computing Architectures

Embedded Systems permeate all aspects of our daily life, from the ubiquitous mobile devices (e.g., PDAs and smart-phones) to play-stations, settop boxes, household appliances, and in every electronic system, be it large or small (e.g., in cars, wrist-watches). Most embedded systems are characterized by stringent design constraints such as reduced memory and computing capacity, severe power and energy restrictions, weight and space limitations, most importantly, very short life spans and thus strict design cycles. Reconfiguration has emerged as a key technology for embedded systems as it offers the promise of increased system performance and component number reduction. Reconfigurable components can be customized or specialized (even dynamically) to the task at hand, thereby executing specific tasks more efficiently leading to possible reductions of the weight and power. In this article, we introduce and discuss compilation techniques for reconfigurable embedded systems. We present specific compiler techniques focusing on source-level code transformations highlighting their potential and the applicability of generative programming techniques to this compilation domain.