Please use this identifier to cite or link to this item:
|Title:||Runtime adaptive extensible embedded processors - A survey|
|Source:||Huynh, H.P.,Mitra, T. (2009). Runtime adaptive extensible embedded processors - A survey. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 5657 LNCS : 215-225. ScholarBank@NUS Repository. https://doi.org/10.1007/978-3-642-03138-0_23|
|Abstract:||Current generation embedded applications demand the computation engine to offer high performance similar to custom hardware circuits while preserving the flexibility of software solutions. Customizable and extensible embedded processors, where the processor core can be enhanced with application-specific instructions, provide a potential solution to this conflicting requirements of performance and flexibility. However, due to the limited area available for implementation of custom instructions in the datapath of the processor core, we may not be able to exploit all custom instruction enhancements of an application. Moreover, a static extensible processor is fundamentally at odds with highly dynamic applications where the custom instructions requirements vary substantially at runtime. In this context, a runtime adaptive extensible processor that can quickly morph its custom instructions and the corresponding custom functional units at runtime depending on workload characteristics is a promising solution. In this article, we provide a detailed survey of the contemporary architectures that offer such dynamic instruction-set support and discuss compiler and/or runtime techniques to exploit such architectures. © 2009 Springer Berlin Heidelberg.|
|Source Title:||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Dec 11, 2017
checked on Dec 9, 2017
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.