Presentation
LRSDP: Low-Rank SDP for Triple Patterning Lithography Layout Decomposition
DescriptionIn this paper, we develop LRSDP, a decomposition-based SDP solver that aims to tackle large-scale TPLD problems with high efficiency. In particular, the positive semidefinite variable along with the diagonal constraints can be decomposed into a variable on a smooth manifold. Then, we apply the augmented Lagrangian method (ALM) to deal with the nonlinear optimization problem and leverage the adaptive regularized semismooth Newton method to search for the inexactly feasible points. Extensive numerical experiments on complex layouts demonstrate the effectiveness of our proposed LRSDP solver compared to other state-of-the-art methods.
Event Type
Research Manuscript
TimeThursday, July 13th11:10am - 11:25am PDT
Location3002, 3rd Floor
EDA
Design for Manufacturing and Reliability