Reconfigurable Computing: Architectures, Tools and Applications: 5th International Workshop, ARC 2009, Karlsruhe, Germany, March 16-18, 2009, Proceedings (Lecture Notes in Computer Science, 5453)

Recon?gurable computing (RC) technologies o?er the promise of substantial performance gains over traditional architectures by customizing, sometimes at run-time, the topology of the underlying architecture to match the speci?c needs of a given application. Contemporary con?gurable architectures allow for the de?nition of architectures with functional and storage units that match the s- ci?c needs of a given computation, in terms of function, bit-width and control structures. Compared to standard microprocessor architectures, advantages are possible in terms of power consumption on a broad range of di?erent application ?elds. Moreover, the ?exibility enabled by recon?guration is also seen as a basic technique for overcoming transient failures in emerging device structures. Techniques for achieving recon?gurable systems are numerous and require the joint development of recon?gurable hardware systems to support the dynamic behavior, e.g., suitable programming models, tools and languages, to support the recon?guration process during run-time as well as during design-time. This includes veri?cation techniques that can demonstrate formally correct recon?- ration sequences at each stage. While there are many problems, the existence and development of technologies such as recent multi- and many-core processor arc- tectures, dynamically recon?gurable and multi-grain computing architectures, as well as application-speci?c processors suggest that there is a very strong need for adaptive and recon?gurable systems.