ELF: Efficient Logic Synthesis by Pruning Redundancy in Refactoring

Tsaras, Dimitris, Li, Xing, Chen, Lei, Xie, Zhiyao, Yuan, Mingxuan

arXiv.org Artificial Intelligence 

--In electronic design automation, logic optimization operators play a crucial role in minimizing the gate count of logic circuits. However, their computation demands are high. Operators such as refactor conventionally form iterative cuts for each node, striving for a more compact representation - a task which often fails 98% on average. Prior research has sought to mitigate computational cost through paralleliza-tion. In contrast, our approach leverages a classifier to prune unsuccessful cuts preemptively, thus eliminating unnecessary resynthesis operations. Experiments on the refactor operator using the EPFL benchmark suite and 10 large industrial designs demonstrate that this technique can speedup logic optimization by 3. 9 on average compared with the state-of-the-art ABC implementation. In logic synthesis, Boolean resynthesis is a technology-independent process that optimizes a logic network to reduce its nodes and levels [1]-[7]. It analyzes the Boolean expressions and iteratively applies local transformations (e.g., operators) to the logic gates and their interconnections. While Boolean resynthesis operators are essential for circuit optimization, they must be used with caution due to their computational expenses. Since most resynthesis operators are heuristic and suboptimal, iterative runs are necessary to identify further potential improvements in the logic circuit [1]. Moreover, the demand for a shorter turn-around time keeps increasing in the fast-paced electronics industry.