TheTo --- A Fast, Scalable and High-Quality Partitioning Driven Placement Tool.

Thumbnail Image

View/Download File

Persistent link to this item

View Statistics

Journal Title

Journal ISSN

Volume Title


TheTo --- A Fast, Scalable and High-Quality Partitioning Driven Placement Tool.

Published Date






Partitioning driven placement approaches are often preferred for fast and scalable solutions to large placement problems. However, due to the inaccuracy of representing wirelength objective by cut objective the quality of such placements often trails the quality of placements produced by pure wirelength driven placements. In this paper we present TheTo, a new partitioning driven placement algorithm that retains the speed associated with traditional partitioning driven placement algorithms but incorporates a number of novel ideas that allows it to produce solutions whose quality is better than those produced by more sophisticated and computationally expensive algorithms. The keys to TheTo's success are a new terminal propagation method that allows the partitioner to exactly map the half-perimeter wirelength cost to min-cut cost, and an integral post-bisectioning refinement step that enhances the effectiveness of the new terminal propagation. Experimental validations show that TheTo is fast and also quite effective in reducing half-perimeter wirelength.



Related to



Series/Report Number

Funding information

Isbn identifier

Doi identifier

Previously Published Citation

Suggested citation

Selvakkumaran, Navaratnasothie; Karypis, George. (2004). TheTo --- A Fast, Scalable and High-Quality Partitioning Driven Placement Tool.. Retrieved from the University Digital Conservancy,

Content distributed via the University Digital Conservancy may be subject to additional license and use restrictions applied by the depositor. By using these files, users agree to the Terms of Use. Materials in the UDC may contain content that is disturbing and/or harmful. For more information, please see our statement on harmful content in digital repositories.