Title :
A Clustering ILP Model for Fast Instruction Selection in Embedded Applicated Specific Processor Design
Author :
Zhao, Kang ; Bian, Jinian
Author_Institution :
Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing
Abstract :
Recently to face the challenges of high efficiency and time-to-market pressure, instruction set auto-customization for application-specific instruction-set processors (ASIP) has become an attractive technology. In this process how to select an optimized instruction set in a short time is crucial to the whole synthesis of ASIP. To address this issue, we formulate the instruction set selection problem with a clustering integer linear programming (ILP) model under multi-constrains, and present a fast cutting plane algorithm to resolve this problem. Finally, the proposed model and algorithm are evaluated through a case study on FIR filter applications
Keywords :
application specific integrated circuits; embedded systems; instruction sets; integer programming; linear programming; microprocessor chips; FIR filter; application-specific instruction-set processors; clustering integer linear programming model; cutting plane algorithm; embedded applicated specific processor design; fast instruction selection; instruction set auto-customization; instruction set selection problem; Application specific processors; Clustering algorithms; Computer science; Electronic design automation and methodology; Finite impulse response filter; Heuristic algorithms; Integer linear programming; Process design; Space exploration; Time to market; ASIP; ILP; Instruction selection; cutting plane algorithm;
Conference_Titel :
Circuits and Systems, 2006. APCCAS 2006. IEEE Asia Pacific Conference on
Conference_Location :
Singapore
Print_ISBN :
1-4244-0387-1
DOI :
10.1109/APCCAS.2006.342347
