AC-Refiner: Efficient Arithmetic Circuit Optimization Using Conditional Diffusion Models

2025-07-03

AI Summary

Existing design-space exploration methods for arithmetic circuits (e.g., adders, multipliers) suffer from low efficiency and struggle to jointly optimize performance, power, and area (PPA). Method: This paper proposes the first conditional diffusion model–based circuit generation framework. It unifies logic synthesis and physical implementation into a quality-of-result (QoR)–guided image-to-circuit generation task. Key innovations include QoR metric–conditioned encoding, Pareto-frontier–driven denoising, and iterative model fine-tuning—enabling end-to-end generation of layout-ready circuit structures from noise. Contribution/Results: Experiments demonstrate substantial improvements in PPA Pareto optimality over state-of-the-art methods. The framework has been successfully integrated into industrial EDA flows, validating its practical deployability and measurable QoR gains.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Probabilistic Circuits and Graphical Models

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Arithmetic circuits, such as adders and multipliers, are fundamental components of digital systems, directly impacting the performance, power efficiency, and area footprint. However, optimizing these circuits remains challenging due to the vast design space and complex physical constraints. While recent deep learning-based approaches have shown promise, they struggle to consistently explore high-potential design variants, limiting their optimization efficiency. To address this challenge, we propose AC-Refiner, a novel arithmetic circuit optimization framework leveraging conditional diffusion models. Our key insight is to reframe arithmetic circuit synthesis as a conditional image generation task. By carefully conditioning the denoising diffusion process on target quality-of-results (QoRs), AC-Refiner consistently produces high-quality circuit designs. Furthermore, the explored designs are used to fine-tune the diffusion model, which focuses the exploration near the Pareto frontier. Experimental results demonstrate that AC-Refiner generates designs with superior Pareto optimality, outperforming state-of-the-art baselines. The performance gain is further validated by integrating AC-Refiner into practical applications.

Problem

• Optimizing arithmetic circuits for performance and efficiency
• Overcoming vast design space and complex constraints
• Improving consistency in exploring high-potential circuit designs

Innovation

• Uses conditional diffusion models for circuit optimization
• Frames circuit synthesis as image generation
• Fine-tunes model with Pareto frontier designs