LMM-IR: Large-Scale Netlist-Aware Multimodal Framework for Static IR-Drop Prediction

2025-11-16

AI Summary

Static IR-drop analysis in chip design is time-consuming and requires frequent iterations, significantly impeding design closure. To address this, we propose the first netlist-aware multimodal IR-drop prediction framework: it models large-scale netlist topologies as 3D point clouds for the first time and introduces a Large Netlist Transformer (LNT) to efficiently process million-node netlists; further, it fuses point-cloud, layout-image, and SPICE-derived features for joint multimodal representation learning in latent space. Evaluated on the ICCAD 2023 contest and against state-of-the-art methods, our approach achieves the highest F1 score and lowest MAE, demonstrating substantial improvements in both accuracy and inference efficiency. This work establishes a scalable, netlist-driven paradigm for high-throughput, large-scale IC IR-drop analysis.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Security and Privacy: Large-scale security measurements
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Static IR drop analysis is a fundamental and critical task in the field of chip design. Nevertheless, this process can be quite time-consuming, potentially requiring several hours. Moreover, addressing IR drop violations frequently demands iterative analysis, thereby causing the computational burden. Therefore, fast and accurate IR drop prediction is vital for reducing the overall time invested in chip design. In this paper, we firstly propose a novel multimodal approach that efficiently processes SPICE files through large-scale netlist transformer (LNT). Our key innovation is representing and processing netlist topology as 3D point cloud representations, enabling efficient handling of netlist with up to hundreds of thousands to millions nodes. All types of data, including netlist files and image data, are encoded into latent space as features and fed into the model for static voltage drop prediction. This enables the integration of data from multiple modalities for complementary predictions. Experimental results demonstrate that our proposed algorithm can achieve the best F1 score and the lowest MAE among the winning teams of the ICCAD 2023 contest and the state-of-the-art algorithms.

Problem

• Predicts static IR drop in chip design using multimodal data integration
• Handles large-scale netlists with up to millions of nodes efficiently
• Reduces computational burden of time-consuming iterative IR drop analysis

Innovation

• Uses large-scale netlist transformer for SPICE processing
• Represents netlist topology as 3D point cloud
• Integrates multimodal data for voltage prediction