DocumentCode :
3276198
Title :
A simple yet effective genetic approach for the orientation assignment on cell-based layout
Author :
Yan, Jin-Tai
Author_Institution :
Dept. of Comput. & Inf. Sci., Nat. Chiao Tung Univ., Hsinchu, Taiwan
fYear :
1996
fDate :
3-6 Jan 1996
Firstpage :
33
Lastpage :
36
Abstract :
In this paper, a genetic approach for the orientation assignment is proposed to minimize total wire length on one cell-based layout. In general, the placement phase in physical design places a set of cells with minimum area or total wire length. It is assumed that all the pins on the cells have been assigned on the fixed positions after the placement phase. In order to reduce the routing area, another advanced improvement will be done to minimize total wire length by assigning the orientation states of cells. In the proposed genetic approach, for any fixed layout, we represent each cell on the layout by a vertical orientation bit and a horizontal orientation bit. Furthermore, the genetic approach assigns the feasible vertical and horizontal states to minimize total wire length. Finally, the proposed approach has been implemented and several benchmarks tested for the orientation assignment. The experimental results show that the proposed genetic approach is effective on these tested benchmarks
Keywords :
VLSI; circuit layout CAD; genetic algorithms; integrated circuit layout; network routing; VLSI design; cell-based layout; genetic approach; horizontal orientation bit; orientation assignment; orientation states; placement phase; routing area reduction; total wire length minimisation; vertical orientation bit; Benchmark testing; Computational modeling; Computer networks; Cost function; Genetics; Information science; Pins; Routing; Simulated annealing; Wire;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
VLSI Design, 1996. Proceedings., Ninth International Conference on
Conference_Location :
Bangalore
ISSN :
1063-9667
Print_ISBN :
0-8186-7228-5
Type :
conf
DOI :
10.1109/ICVD.1996.489450
Filename :
489450
Link To Document :
بازگشت