ReSPack: A Large-Scale Rectilinear Steiner Tree Packing Benchmark for Wire Routing

Kanghoon Lee, Young‐Joon Park, Han-Seul Jeong, D.j. Yoon, Sunghoon Hong, Sungryull Sohn, Minu Kim, Hanbum Ko, Moontae Lee, Honglak Lee, Kyunghoon Kim, Euihyuk Kim, Seonggeon Cho, Jaesang Min, Woohyung LimZenodo 2022 · DOI · OpenAlex

Abstract

Combinatorial optimization (CO) has been studied as a useful tool for modeling industrial problems, but it still remains a challenge in complex domains because of the NP-hardness. With recent advances in machine learning, the field of CO is shifting to the study of neural combinatorial optimization using a large amount of data, showing promising results in some CO problems. Rectilinear Steiner Tree Packing Problem (RSTPP) is a well-known CO problem and is widely used in modeling wiring problem among components in a printed circuit board and an integrated circuit design. Despite the importance for a design automation, the lack of availab le data has restricted to fully leverage machine learning approaches. In this paper, we present ReSPack, a large-scale synthetic RSTPP dataset and a benchmark for wire routing. ReSPack includes a source code for generating RSTPP problems of various types with different sizes, data instances generated for the benchmark testing, and implementations of several baseline algorithms. We believe that ReSPack can contribute to solving circuit design problems as well as the development of CO with machine learning.
BibTeX
@inproceedings{lee2022respack,
  author = {Kanghoon Lee and Young‐Joon Park and Han-Seul Jeong and D.J. Yoon and Sunghoon Hong and Sungryull Sohn and Minu Kim and Hanbum Ko and Moontae Lee and Honglak Lee and Kyunghoon Kim and Euihyuk Kim and Seonggeon Cho and Jaesang Min and Woohyung Lim},
  title = {{ReSPack: A Large-Scale Rectilinear Steiner Tree Packing Benchmark for Wire Routing}},
  booktitle = {Zenodo},
  year = {2022},
  doi = {10.5281/zenodo.6961545},
}

Loading related content...

Figure