在積體電路的實體設計流程中，繞線是一個非常複雜且耗時的階段。在完成繞線後，將執行設計規則檢查以檢查電路之正確性及可製造性，並且必須消除發現的違規。然而，違規消除過程需要耗費大量時間，因為對電路的任何修改都必須重新執行繞線。因此，開發一個無需執行繞線階段即可預測違規結果的預測器是非常有幫助的。然而，由於繞線器演算法中的隨機因子，即使是使用同樣的擺置結果做為繞線輸入，每次的繞線結果也可能會有些許不同。因此，預測器需要能夠藉由相同的擺置來預測大部分可能產生的繞線結果。

本論文提出了一種從元件擺置結果自動生成繞線指南的方法，使商業繞線器能夠生成品質更好的繞線結果。每個繞線指南都會更改特定區域中的全域邊容量百分比，從而更改原有繞線行為。我們的方法採用機器學習和啟發式技術來 (1) 從佈局和一組繞線指南中快速預測 DRC 熱點，以及 (2) 有效地使用DRC熱點預測器從元件擺置結果中生成輔助繞線器的繞線指南並使其盡可能減少 DRC熱點的產生。優越的實驗結果顯示出此方法之可行性。

The routing stage is very complex and time-consuming in the physical design flow of modern circuits. After the routing stage, the Design Rule Check (DRC) will be performed to check the circuit correctness and manufacturability, and the found DRC violations (i.e., DRC hotspots) must be eliminated afterward. However, the DRC hotspot elimination process takes lots of time since any modification to the design must iteratively re-perform routing. Hence, developing a predictor that can predict DRC results without performing routing is beneficial. Nevertheless, the non-deterministic behavior in a router may cause different DRC hotspot distributions with an identical placement result. Therefore, the DRC hotspot predictor needs to predict all the possible DRC hotspots generated by different routing results.

This thesis presents an approach to automatically generate routing guides from a placement instance, enabling a commercial router to produce a routing solution of better quality. Each routing guide changes the global edge capacity in a particular area and hence changes the default routing behavior.
Our approach adopts machine learning and heuristic techniques to (1) fast predict DRC hotspots from a placement instance and a set of routing guides and (2) effectively employ the DRC hotspot predictor to generate routing guides that assist the router in producing as few DRC hotspots as possible. Encouraging experimental results are shown to support our approach.

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Previous Work on Congestion Estimation and Prediction . . 3
1.3 Our Contributions . . . . . . . . . . . . . . . . . . . . 4
1.4 Thesis Organization . . . . . . . . . . . . . . . . . . . 5
2 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . 6
2.1 Semantic Segmentation . . . . . . . . . . . . . . . . . . 6
2.2 U-Net . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Smoothening Process . . . . . . . . . . . . . . . . . . . 7
2.4 Object Detection . . . . . . . . . . . . . . . . . . . . . 9
2.5 YOLOv3 . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.5.1 Anchor . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.5.2 Loss Function . . . . . . . . . . . . . . . . . . . . . 12
2.6 Routing Guide . . . . . . . . . . . . . . . . . . . . . . 14
2.7 Problem Description . . . . . . . . . . . . . . . . . . . 14
3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . 15
3.1 Overall Flow . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 Feature Extraction . . . . . . . . . . . . . . . . . . . . 18
3.2.1 Placer-generated Features . . . . . . . . . . . . . . . 18
3.2.2 Horizontal/Vertical Macro-aware RUDY Features . . . . . 19
3.3 Generation of Routing Guides for Training . . . . . . . . 22
3.3.1 Label-based Routing Guide Generation . . . . . . . . . . 22
3.3.2 Random Macro-aware Routing Guide Generation . . . . . . 22
3.4 Image Generation . . . . . . . . . . . . . . . . . . . . . 23
3.5 Labels Generation . . . . . . . . . . . . . . . . . . . . 24
3.6 Data Augmentation . . . . . . . . . . . . . . . . . . . . 24
3.7 Routing Guide Updating . . . . . . . . . . . . . . . . . . 25
4 Experimental Results . . . . . . . . . . . . . . . . . . . . 26
4.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . 26
4.2 Testing Results of Phase-1 Predictors . . . . . . . . . . 28
4.3 Testing Results of Phase-2 Predictors . . . . . . . . . . 33
4.4 Results of DRC Violations with Generated Routing Guides . 37
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 41
References . . . . . . . . . . . . . . . . . . . . . . . . . . 42

[1] “Synopsys, ic compiler ii: Block-level implementation,” 2019.
[2] I. S. Bustany, D. Chinnery, J. R. Shinnerl, and V. Yutsis, “Ispd 2015 benchmarks with fence regions and routing blockages for detailed-routing-driven placement,” in Proceedings of Symposium on International Symposium on Physical Design, p. 157–164, 2015.
[3] R. Hadsell and P. Madden, “Improved global routing through congestion estimation,” in Proceedings of Design Automation Conference, pp. 28–31, 2003.
[4] J. Lou, S. Krishnamoorthy, and H. S. Sheng, “Estimating routing congestion using probabilistic analysis,” in Proceedings of International Symposium on Physical Design, pp. 112–117, 2001.
[5] P. Spindler and F. M. Johannes, “Fast and accurate routing demand estimation for efficient routability-driven placement,” in Proceedings of Design, Automation and Test in Europe, pp. 1226–1231, 2007.
[6] J. Westra, C. Bartels, and P. Groeneveld, “Probabilistic congestion prediction,” in Proceedings of International Symposium on Physical Design, pp. 204–209, 2004.
[7] J. Chen, J. Kuang, G. Zhao, D. J.-H. Huang, and E. F. Young, “Pros: A plug-in for routability optimization applied in the state-of-the-art commercial eda tool using deep learning,” in Proceedings of International Conference on Computer-Aided Design, 2020.
[8] S. Liu, Q. Sun, P. Liao, Y. Lin, and B. Yu, “Global placement with deep learning-enabled explicit routability optimization,” in Proceedings of Design, Automation and Test in Europe, pp. 1821–1824, 2021.
[9] C. Ma, Y. Xiao, S. Wang, J. Yu, and J. Chen, “Congestnn: An bi-directional congestion prediction framework for large-scale heterogeneous fpgas,” in Proceedings of International Conference on ASIC, 2021.
[10] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in Proceedings of Conference on Computer Vision and Pattern Recognition, pp. 3431–3440, 2015.
[11] M. B. Alawieh, W. Li, Y. Lin, L. Singhal, M. A. Iyer, and D. Z. Pan, “High-definition routing congestion prediction for large-scale fpgas,” in Proceedings of Asia and South Pacific Design Automation Conference, pp. 26–31, 2020.
[12] B. Wang, G. Shen, D. Li, J. Hao, W. Liu, Y. Huang, H. Wu, Y. Lin, G. Chen, and P. A. Heng, “Lhnn: Lattice hypergraph neural network for vlsi congestion prediction,” in Proceedings of Design Automation Conference, pp. 1297–1302, 2022.
[13] M. Su, H. Ding, S. Weng, C. Zou, Z. Zhou, Y. Chen, J. Chen, and Y.-W. Chang, “Highcorrelation 3d routability estimation for congestion-guided global routing,” in Proceedings of Asia and South Pacific Design Automation Conference, pp. 580–585, 2022.
[14] Y. Pan, Z. Zhou, and A. Ivanov, “Routability-driven global routing with 3d congestion estimation using a customized neural network,” in Proceedings of International Symposium on Quality Electronic Design, 2022.
[15] J. Liu, C.-W. Pui, F. Wang, and E. F. Y. Young, “Cugr: Detailed-routability-driven 3d global routing with probabilistic resource model,” in Proceedings of Design Automation Conference, 2020.
[16] W.-T. Hung, J.-Y. Huang, Y.-C. Chou, C.-H. Tsai, and M. Chao, “Transforming global routing report into drc violation map with convolutional neural network,” in Proceedings of International Symposium on Physical Design, pp. 57–64, 2020.
[17] L. Li, Y. Cai, and Q. Zhou, “An efficient approach for drc hotspot prediction with convolutional neural network,” in Proceedings of International Symposium on Circuits and Systems, 2021.
[18] W. Zeng, A. Davoodi, and R. O. Topaloglu, “Explainable drc hotspot prediction with random forest and shap tree explainer,” in Proceedings of Design, Automation and Test in Europe, pp. 1151–1156, 2020.
[19] R. Chen, W. Zhong, H. Yang, H. Geng, X. Zeng, and B. Yu, “Faster region-based hotspot detection,” in Proceedings of Design Automation Conference, 2019.
[20] R. Liang, H. Xiang, D. Pandey, L. Reddy, S. Ramji, G.-J. Nam, and J. Hu, “Drc hotspot prediction at sub-10nm process nodes using customized convolutional network,” in Proceedings of International Symposium on Physical Design, pp. 135–142, 2020.
[21] Z. Xie, Y.-H. Huang, G.-Q. Fang, H. Ren, S.-Y. Fang, Y. Chen, and J. Hu, “Routenet: Routability prediction for mixed-size designs using convolutional neural network,” in Proceedings of International Conference on Computer-Aided Design, 2018.
[22] A. Ghose, V. Zhang, Y. Zhang, D. Li, W. Liu, and M. Coates, “Generalizable cross-graph embedding for gnn-based congestion prediction,” in Proceedings of International Conference on Computer-Aided Design, 2021.
[23] R. Kirby, S. Godil, R. Roy, and B. Catanzaro, “Congestionnet: Routing congestion prediction using deep graph neural networks,” in Proceedings of International Conference on Very Large Scale Integration, pp. 217–222, 2019.
[24] X. Chen, Z. Di, W. Wu, Q. Wu, J. Shi, and Q. Feng, “Detailed routing short violation prediction using graph-based deep learning model,” in IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 69, no. 2, pp. 564–568, 2021.
[25] C.-C. Chang, J. Pan, T. Zhang, Z. Xie, J. Hu, W. Qi, C.-W. Lin, R. Liang, J. Mitra, E. Fallon, et al., “Automatic routability predictor development using neural architecture search,” in Proceedings of International Conference on Computer-Aided Design, 2021.
[26] H.-H. Pan, “Macro-aware drc hotspot prediction based on deep learning,” Master’s thesis, National Tsing Hua University, 2021.
[27] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in Proceedings of International Conference on Medical image computing and computer-assisted intervention, pp. 234–241, 2015.
[28] J. Redmon and A. Farhadi, “Yolov3: An incremental improvement,” arXiv preprint arXiv:1804.02767, 2018.
[29] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of IEEE conference on Computer Vision and Pattern Recognition, pp. 770–778, 2016.
[30] T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in Proceedings of IEEE conference on Computer Vision and Pattern Recognition, pp. 2117–2125, 2017.
[31] H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, and S. Savarese, “Generalized intersection over union: A metric and a loss for bounding box regression,” in Proceedings of IEEE conference on Computer Vision and Pattern Recognition, pp. 658–666, 2019.
[32] Z. Zheng, P. Wang, W. Liu, J. Li, R. Ye, and D. Ren, “Distance-iou loss: Faster and better learning for bounding box regression,” in Proceedings of AAAI conference on artificial intelligence, vol. 34, pp. 12993–13000, 2020.
[33] T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, “Focal loss for dense object detection,” in Proceedings of IEEE International Conference on Computer Vision, pp. 2980–2988, 2017.
[34] M. Ester, H.-P. Kriegel, J. Sander, and X. Xu, “A density-based algorithm for discovering clusters in large spatial databases with noise,” in Proceedings of Second International Conference on Knowledge Discovery and Data Mining, pp. 226–231, 1996.
[35] A. Caldwell, A. Kahng, S. Mantik, I. Markov, and A. Zelikovsky, “On wirelength estimations for row-based placement,” in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 18, no. 9, pp. 1265–1278, 1999.