Abstract
The ability to accurately and efficiently estimate the routability of a circuit based on its placement is one of the most challenging and difficult tasks in the Field Programmable Gate Array (FPGA) flow. In this article, we present a novel, deep learning framework based on a Convolutional Neural Network (CNN) model for predicting the routability of a placement. Since the performance of the CNN model is strongly dependent on the hyper-parameters selected for the model, we perform an exhaustive parameter tuning that significantly improves the model’s performance and we also avoid overfitting the model. We also incorporate the deep learning model into a state-of-the-art placement tool and show how the model can be used to (1) avoid costly, but futile, place-and-route iterations, and (2) improve the placer’s ability to produce routable placements for hard-to-route circuits using feedback based on routability estimates generated by the proposed model. The model is trained and evaluated using over 26K placement images derived from 372 benchmarks supplied by Xilinx Inc. We also explore several opportunities to further improve the reliability of the predictions made by the proposed DLRoute technique by splitting the model into two separate deep learning models for (a) global and (b) detailed placement during the optimization process. Experimental results show that the proposed framework achieves a routability prediction accuracy of 97% while exhibiting runtimes of only a few milliseconds.
- Z. Abuowaimer, D. Maarouf, T. Martin, J. Foxcroft, G. Grewal, S. Areibi, and A. Vannelli. 2018. GPLace3.0: Routability-driven analytic placer for ultrascale FPGA architectures. ACM Trans. Des. Autom. Electron. Syst. 23, 5 (Oct. 2018), 66:1–66:33. Google Scholar
Digital Library
- A. Al-Hyari and S. Areibi. 2017. Design space exploration of convolutional neural networks based on evolutionary algorithms. J. Computat. Vis. Imag. Syst. 3, 1 (Oct. 2017), 1–3.Google Scholar
- A. Alhyari, A. Shamli, Z. Abuowaimer, G. Grewal, and S. Areibi. 2019. A deep learning framework to predict routability for FPGA circuit placement. In International Conference on Field Programmable Logic and Applications. 1–8.Google Scholar
- S. Brown, J. Rose, and Z. Vranesic. 1993. A stochastic model to predict the routability of field programmable gate arrays. IEEE Trans. Comput.-aided Des. Integ. Circ. Syst. 12, 12 (Dec. 1993), 1827–1838. Google Scholar
Digital Library
- P. Chan, M. Schlag, and J. Zienr. 1993. On routability prediction for field programmable gate arrays. In Design Automation Conference. 326–330. Google Scholar
Digital Library
- W. Chan, P. Ho, A. Kahng, and P. Saxena. 2017. Routability optimization for industrial designs at Sub-14nm process nodes using machine learning. In International Symposium on Physical Design. 15–21. Google Scholar
Digital Library
- W. T. J. Chan, Y. Du, A. B. Kahng, S. Nath, and K. Samadi. 2016. BEOL stack-aware routability prediction from placement using data mining techniques. In International Conference on Computer Design (ICCD’16). 41–48.Google Scholar
- B. Debowski, S. Areibi, G. Grewal, and J. Tempelman. 2012. A dynamic sampling framework for multi-class imbalanced data. In International Conference on Machine Learning and Applications. 113–118. Google Scholar
Digital Library
- I. Goodfellow, Y. Bengio, and A. Courville. 2015. Deep Learning. The MIT Press, Cambridge, MA. Google Scholar
Digital Library
- G. Grewal and S. Areibi. 2018. Guelph FPGA CAD Group. Retrieved from http://fpga.socs.uoguelph.ca/.Google Scholar
- A. Gulli and S. Pal. 2017. Deep Learning with Keras. Packt Publishing Ltd. Google Scholar
Digital Library
- PariVallal Kannan, Shankar Balachandran, and Dinesh Bhatia. 2001. fGREP - Fast generic routing demand estimation for placed FPGA circuits. In International Conference on Field-programmable Logic and Applications. Springer-Verlag, 37–47. Google Scholar
Digital Library
- P. Kannan, S. Calachandran, and D. Bhatia. 2004. On metrics for comparing interconnect estimation methods for FPGAs. IEEE Trans. VLSI 12, 4 (Apr. 2004), 381–385. Google Scholar
Digital Library
- Guillaume Lemaître, Fernando Nogueira, and Christos K. Aridas. 2017. Imbalanced-learn: A Python toolbox to tackle the curse of imbalanced datasets in machine learning. J. Mach. Learn. Res. 18, 17 (2017), 1–5. Retrieved from http://jmlr.org/papers/v18/16-365.html. Google Scholar
Digital Library
- W. Li, S. Dhar, and D. Pan. 2018. UTPlaceF: A routability-driven FPGA placer with physical and congestion aware packing. IEEE Trans. CAD Syst. 37, 4 (2018), 869–882. Google Scholar
Digital Library
- Jinan Lou, Shashidhar Thakur, Shankar Krishnamoorthy, and Henry S. Sheng. 2002. Estimating routing congestion using probabilistic analysis. IEEE Trans. Comput.-aided Des. ICs Syst. 21, 2 (Jan. 2002).Google Scholar
- D. Maarouf, A. Alhyari, Z. Abuowaimer, T. Martin, A. Gunter, G. Grewal, S. Areibi, and A. Vannelli. 2018. A machine-learning based congestion estimation for modern FPGAs. In IEEE International Conference on Field-programmable Logic and Applications. 427–434.Google Scholar
- D. Marrouff, A. Shamli, T. Martin, G. Grewal, and S. Areibi. 2020. A deep-learning framework for predicting congestion during FPGA placement. In 30th International Conference on Field-programmable Logic and Applications. 138–144.Google Scholar
- F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. 2011. Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 12 (2011), 2825–2830. Google Scholar
Digital Library
- C. Pui, G. Chen, W. Chow, K. Lam, P. Tu, H. Zhang, E. Young, and B. Yu. 2016. RippleFPGA: A routability-driven placement for large-scale heterogeneous FPGAs. In International Conference on Computer-aided Design. 1–8. Google Scholar
Digital Library
- C. Pui, G. Chen, Y. Ma, E. Young, and B. Yu. 2017. Clock-aware ultrascale FPGA placement with machine learning routability prediction. In IInternational Conference on Computer-aided Design. ACM, 929–936. Google Scholar
Digital Library
- Z. Qi, Y. Cai, and Q. Zhou. 2014. Accurate prediction of detailed routing congestion using supervised data learning. In IEEE International Conference on Computer Design. 97–103.Google Scholar
- A. Rahman, S. Oh, J. Lee, and K. Choi. 2017. Design space exploration of FPGA accelerators for convolutional neural networks. In Design, Automation & Test in Europe Conference (DATE’17). 1147–1152. Google Scholar
Digital Library
- A. Tabrizi, N. Darav, L. Rakai, I. Bustany, A. Kennings, and L. Behjat. 2020. Eh? Predictor: A deep learning framework to identify detailed routing short violations from a placed netlist. IEEE Trans. Comput.-aided Des. Integ. Circ. Syst. 39, 6 (June 2020), 1177–1190.Google Scholar
Cross Ref
- R. Glenn Wood and R. Rutenbar. 1997. FPGA Routing and routability estimation via boolean satisfiability. In International Symposium on FPGAs. 119–125. Google Scholar
Digital Library
- Z. Xie, Y. Huang, G. Fang, H. Ren, S. Fang, Y. Chen, and J. Hu. 2018. RouteNet: Routability prediction for mixed-size designs using convolutional neural network. In International Conference on Computer-aided Design. 1–8. Google Scholar
Digital Library
- Xilinx. 2016. ISPD 2016 Routability-Driven FPGA Placement Contest. Retrieved from http://www.ispd.cc/contests/16/ispd2016_contest.html. Google Scholar
Digital Library
- S. Yang, A. Gayasen, C. Mulpuri, S. Reddy, and R. Aggarwal. 2016. Routability-driven FPGA placement contest. In International Symposium on Physical Design. 139–143. Google Scholar
Digital Library
- D. Yeager, D. Chiu, and G. Lemieux. 2007. Congestion estimation and localization in FPGAs: A visual tool for interconnect prediction. In International Workshop on System Level Interconnect Prediction. ACM, 33–40. Google Scholar
Digital Library
- C. Yu and Z. Zhang. 2019. Painting on placement: Forecasting routing congestion using conditional generative adversarial nets. In Design Automation Conference (DAC’19). 1–6. Google Scholar
Digital Library
- J. Zhao, T. Liang, S. Sinha, and W. Zhang. 2019. Machine learning based routing congestion prediction in FPGA high-level synthesis. In Design, Automation and Test in Europe (DATE’19). 1130–1135.Google Scholar
- Q. Zhou, X. Wang, Z. Qi, Z. Chen, Q. Zhou, and Y. Cai. 2015. An accurate detailed routing routability prediction model in placement. In Asia Symposium on Quality Electronic Design. 119–122.Google Scholar
- Y. Zhuo, H. Li, and S. Mohanty. 2006. A congestion driven placement algorithm for FPGA synthesis. In International Conference on Field-programmable Logic and Applications. 683–686.Google Scholar
Index Terms
A Deep Learning Framework to Predict Routability for FPGA Circuit Placement
Recommendations
Novel Congestion-estimation and Routability-prediction Methods based on Machine Learning for Modern FPGAs
Special Section on Security in FPGAs and Regular ArticlesEffectively estimating and managing congestion during placement can save substantial placement and routing runtime. In this article, we present a machine-learning model for accurately and efficiently estimating congestion during FPGA placement. Compared ...
Routability-driven analytical placement for mixed-size circuit designs
ICCAD '11: Proceedings of the 2011 IEEE/ACM International Conference on Computer-Aided DesignDue to the significant mismatch between existing wirelength models and the congestion objective in placement, considering routability during placement is particularly significant for modern circuit designs. In this paper, a novel routability-driven ...
Routability-driven analytical placement for mixed-size circuit designs
ICCAD '11: Proceedings of the International Conference on Computer-Aided DesignDue to the significant mismatch between existing wirelength models and the congestion objective in placement, considering routability during placement is particularly significant for modern circuit designs. In this paper, a novel routability-driven ...






Comments