Title: Routing Congestion Estimation with Real Design Constraints
Authors: Liu, Wen-Hao
Wei, Yaoguang
Sze, Cliff
Alpert, Charles J.
Li, Zhuo
Li, Yih-Lang
Viswanathan, Natarajan
資訊工程學系
Department of Computer Science
Keywords: Congestion estimation;global routing;routability
Issue Date: 2013
Abstract: To address the routability issue, routing congestion estimators (RCE) become essential in industrial design flow. Recently, several RCEs [1-4] based on global routing engines are developed, but they typically ignore the effects of routing on timing so that the identified routing paths may be overlong and thus impractical. To be aware of the timing issues, our proposed global-routing-based RCE obeys the layer directive and scenic constraints to respectively limit the routing layers and the maximum routing wirelength of the potentially timing-critical nets. To handle the scenic constrains, we propose a novel method based on a relaxation-legalization scheme. Also, because the work in [5] reveals that congestion ratio is a better indicator than overflow to evaluate routability, this work focuses on minimizing the congestion ratio rather than overflows. As will be shown, the problem of minimizing congestion ratio is more complicated than minimizing overflows, so we develop a new rip-up and rerouting scheme to reduce congestion and further to approach a target congestion ratio. Moreover, to fit the demands of practical uses, this work presents a control utility to trade off runtime and quality, which is an essential function to an industrial RCE tool. Experiments reveal that the proposed RCE is faster and more accurate than another industrial global-routing-based RCE.
URI: http://hdl.handle.net/11536/23071
ISBN: 978-1-4503-2071-9
ISSN: 0738-100X
Journal: 2013 50TH ACM / EDAC / IEEE DESIGN AUTOMATION CONFERENCE (DAC)
Appears in Collections:Conferences Paper