Cost-generic logic synthesis

Resubstitution with customizable cost functions in the EPFL logic network library mockturtle — 1st place, IWLS 2022 contest.

A cost-generic resubstitution framework that lets users plug arbitrary cost functions (area, depth, fanout, and beyond) into logic optimization (Wang et al., 2022), generalized into the AnySyn framework (Wang et al., 2023). The algorithm ships in mockturtle, the EPFL open-source logic network library, and won 1st place in the IWLS 2022 programming contest.

References

2023

arXiv
AnySyn: A Cost-Generic Logic Synthesis Framework with Customizable Cost Functions

Hanyu Wang, Siang-Yun Lee, and Giovanni De Micheli

arXiv preprint, 2023

arXiv Bib HTML
@article{wang2023anysyn, author = {Wang, Hanyu and Lee, Siang-Yun and De Micheli, Giovanni}, title = {AnySyn: A Cost-Generic Logic Synthesis Framework with Customizable Cost Functions}, journal = {arXiv preprint}, volume = {arXiv:2311.14721}, year = {2023}, }

2022

IWLS
A Cost-Generic Resubstitution Algorithm with Customizable Cost Functions

Hanyu Wang, Siang-Yun Lee, and Giovanni De Micheli

In International Workshop on Logic & Synthesis (IWLS), 2022

Page Bib HTML
@inproceedings{wang2022costgeneric, author = {Wang, Hanyu and Lee, Siang-Yun and De Micheli, Giovanni}, title = {A Cost-Generic Resubstitution Algorithm with Customizable Cost Functions}, booktitle = {International Workshop on Logic \& Synthesis (IWLS)}, year = {2022}, }