用物理引导的扩散模型生成更优芯片宏单元布局
Physics-Guided Geometric Diffusion for Macro Placement Generation
- 双域去噪架构融合图神经网络与注意力机制
- 在ISPD2005数据集上降低6.1%-6.2%布线长度
- 大尺寸设计下稳定收敛，适合工业级芯片布局
宏单元布局是超大规模集成电路物理设计的关键环节，直接影响芯片性能。现有数据驱动方法在处理序列依赖关系及平衡拓扑连通性与物理约束方面存在不足。为此，我们提出MacroDiff+，一种物理引导的几何扩散框架。其设计双域去噪结构，通过异质图神经网络编码拓扑连通性，利用Transformer建模全局几何上下文。同时引入物理引导采样策略，通过显式梯度主动引导生成过程，确保结果兼具统计合理性与物理可实现性。在ISPD2005 MMS基准测试中，MacroDiff+相比最优基线降低6.1%-6.2%布线长度，且在大规模设计中表现出更强稳定性与可扩展性，而此前方法常无法收敛。代码已开源：https://github.com/jhy00n/MacroDiff-plus。
原文摘要 · Abstract (English)
Macro placement is a pivotal stage in VLSI physical design, fundamentally determining the overall chip performance. Recent data-driven placement methods have demonstrated significant potential, yet they often struggle to handle sequential dependencies and to balance topological connectivity with physical constraints. To bridge this gap, we propose MacroDiff+, a physics-guided geometric diffusion framework. Specifically, we design a dual-domain denoising architecture that couples topological connectivity encoded by heterogeneous GNNs with global geometric context modeled by a Transformer. Furthermore, we introduce Physics-Guided Sampling, an inference strategy that actively steers the generation using explicit gradients to ensure both statistical plausibility and physical validity. On the ISPD2005 MMS benchmarks, MacroDiff+ outperforms state-of-the-art baselines with a 6.1-6.2% reduction in wirelength. Notably, it exhibits superior stability and scalability on large-scale designs where prior methods fail to converge. The source code is available at https://github.com/jhy00n/MacroDiff-plus.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。