Litcius/Paper detail

AlphaSyn: Logic Synthesis Optimization with Efficient Monte Carlo Tree Search

Zehua Pei, Fangzhou Liu, Zhuolun He, Guojin Chen, Haisheng Zheng, Keren Zhu, Bei Yu

202318 citationsDOI

Abstract

Recent years have seen rising research in logic synthesis recipe generation to improve the Quality-of-Result (QoR). However, existing approaches typically have low efficiency and are stuck at local optima. In this work, we propose a logic synthesis optimization framework, AlphaSyn, that incorporates a domain-specific Monte Carlo tree search (MCTS) algorithm. AlphaSyn enables exploration across the entire search space while optimizing sampling points utilization. We further develop a synthesis-specific upper confidence bound for trees (SynUCT) algorithm for the selection phase and a well-designed learning strategy to enhance the stability of the MCTS algorithm. The AlphaSyn algorithm is fully parallelized for efficiency with asynchronous MCTS exploration and significance-base resource allocation. For standard-cell technology mapping on the ASAP 7nm library among other tasks, experimental results show that AlphaSyn outperforms SOTA FlowTune with an average 8.74% area reduction and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{1.24}\times$</tex> runtime speedup.

Topics & Concepts

Monte Carlo tree searchComputer scienceSpeedupAsynchronous communicationTree (set theory)Monte Carlo methodDomain (mathematical analysis)AlgorithmReduction (mathematics)Theoretical computer scienceMathematical optimizationParallel computingMathematicsMathematical analysisComputer networkGeometryStatisticsSoftware Engineering ResearchMachine Learning and AlgorithmsMachine Learning in Materials Science
AlphaSyn: Logic Synthesis Optimization with Efficient Monte Carlo Tree Search | Litcius