DiffPattern-Flex: Efficient Layout Pattern Generation via Discrete Diffusion

2025-05-07

AI Summary

Neural network–based placement generation in chip physical design often fails to guarantee layout legality, necessitating costly post-processing. Method: This paper proposes the first placement generation framework integrating discrete diffusion modeling with white-box, rule-driven optimization. It employs a lightweight, lossless layout representation, tightly couples real-time design rule checking with efficient sampling, and unifies generative synthesis and legalization. Crucially, it embeds interpretable, hard design constraints directly into the discrete diffusion process—eliminating the need for iterative correction. Contribution/Results: Experiments across multiple benchmarks demonstrate 100% compliance with critical design rules (e.g., spacing, density, connectivity), significantly faster generation than state-of-the-art methods, and simultaneous preservation of topological diversity and structural robustness.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Mixed Discrete/Continuous Search
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Recent advancements in layout pattern generation have been dominated by deep generative models. However, relying solely on neural networks for legality guarantees raises concerns in many practical applications. In this paper, we present ool{DiffPattern}-Flex, a novel approach designed to generate reliable layout patterns efficiently. ool{DiffPattern}-Flex incorporates a new method for generating diverse topologies using a discrete diffusion model while maintaining a lossless and compute-efficient layout representation. To ensure legal pattern generation, we employ {an} optimization-based, white-box pattern assessment process based on specific design rules. Furthermore, fast sampling and efficient legalization technologies are employed to accelerate the generation process. Experimental results across various benchmarks demonstrate that ool{DiffPattern}-Flex significantly outperforms existing methods and excels at producing reliable layout patterns.

Problem

• Generate reliable layout patterns efficiently
• Ensure legal pattern generation with design rules
• Accelerate generation with fast sampling techniques

Innovation

• Uses discrete diffusion model for diverse topologies
• Employs optimization-based white-box pattern assessment
• Integrates fast sampling and efficient legalization technologies