DocumentCode :
1370367
Title :
Faster minimization of linear wirelength for global placement
Author :
Alpert, Charles J. ; Chan, Tony F. ; Kahng, Andrew B. ; Markov, Igor L. ; Mulet, Pep
Author_Institution :
Res. Lab., IBM Corp., Austin, TX, USA
Volume :
17
Issue :
1
fYear :
1998
fDate :
1/1/1998 12:00:00 AM
Firstpage :
3
Lastpage :
13
Abstract :
A linear wirelength objective more effectively captures timing, congestion, and other global placement considerations than a squared wirelength objective. The GORDIAN-L cell placement tool minimizes linear wirelength by first approximating the linear wirelength objective by a modified squared wirelength objective, then executing the following loop-(1) minimize the current objective to yield some approximate solution and (2) use the resulting solution to construct a more accurate objective-until the solution converges. This paper shows how to apply a generalization of an algorithm due to Weiszfeld (1937) to placement with a linear wirelength objective and that the main GORDIAN-L loop is actually a special case of this algorithm. We then propose applying a regularization parameter to the generalized Weiszfeld algorithm to control the tradeoff between convergence and solution accuracy; the GORDIAN-L iteration is equivalent to setting this regularization parameter to zero. We also apply novel numerical methods, such as the primal-Newton and primal-dual Newton iterations, to optimize the linear wirelength objective. Finally, we show both theoretically and empirically that the primal-dual Newton iteration stably attains quadratic convergence, while the generalized Weiszfeld iteration is linear convergent. Hence, primal-dual Newton is a superior choice for implementing a placer such as GORDIAN-L, or for any linear wirelength optimization
Keywords :
Newton method; circuit layout CAD; convergence of numerical methods; integrated circuit layout; iterative methods; minimisation; GORDIAN-L cell placement tool; generalized Weiszfeld algorithm; global placement; linear convergence; linear wirelength minimisation; linear wirelength objective; numerical methods; primal-Newton iteration; primal-dual Newton iteration; quadratic convergence; regularization parameter; solution accuracy; timing; Associate members; Delay lines; Equations; Gravity; Helium; Integrated circuit layout; Linear approximation; Optimization methods; Partitioning algorithms; Timing;
fLanguage :
English
Journal_Title :
Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions on
Publisher :
ieee
ISSN :
0278-0070
Type :
jour
DOI :
10.1109/43.673628
Filename :
673628
Link To Document :
بازگشت