DocumentCode :
1153071
Title :
Automated custom instruction generation for domain-specific processor acceleration
Author :
Clark, Nathan T. ; Zhong, Hongtao ; Mahlke, Scott A.
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Michigan Univ., Ann Arbor, MI, USA
Volume :
54
Issue :
10
fYear :
2005
Firstpage :
1258
Lastpage :
1270
Abstract :
Application-specific extensions to the computational capabilities of a processor provide an efficient mechanism to meet the growing performance and power demands of embedded applications. Hardware, in the form of new function units (or coprocessors), and the corresponding instructions are added to a baseline processor to meet the critical computational demands of a target application. In this paper, the design of a system to automate the instruction set customization process is presented. A dataflow graph design space exploration engine efficiently identifies computation subgraphs to create custom hardware and a compiler subgraph matching framework seamlessly exploits this hardware. We demonstrate the effectiveness of this system across a range of application domains and study the applicability of the custom hardware across an entire application domain. Generalization techniques are presented which enable the application-specific hardware to be more effectively used across a domain.
Keywords :
data flow computing; data flow graphs; embedded systems; instruction sets; multiprocessing systems; program compilers; automated custom instruction generation; baseline processor; compiler subgraph matching framework; dataflow graph design space exploration engine; domain-specific processor acceleration; embedded application; instruction set design; Acceleration; Application specific integrated circuits; Computer aided instruction; Coprocessors; Costs; Embedded computing; Engines; Hardware; Power demand; Space exploration; Index Terms- Automatic synthesis; instruction set design; instruction set interpretation; special-purpose; special-purpose and application-based systems.;
fLanguage :
English
Journal_Title :
Computers, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9340
Type :
jour
DOI :
10.1109/TC.2005.156
Filename :
1501791
Link To Document :
بازگشت