MAGE: Human-Like Macro Placement via Agentic Multimodal Reasoning

2026-07-20

AI Summary

This work addresses the heavy reliance on manual intervention in macro placement for industrial physical design, where automated methods lack human-like structural awareness. The authors propose MAGE, a framework that decomposes placement into a six-stage workflow integrating structured floorplanning rules, visual inspection, and iterative refinement. Expert knowledge is injected via natural language instructions, and high-quality feedback is propagated through a multi-agent bidding mechanism. For the first time, the method introduces unsupervised, human-inspired four-dimensional layout quality metrics—gap, void, pocket, and alignment scores—enabling a fully unsupervised optimization paradigm. Evaluated on nine designs in NanGate45 and GF12nm technologies, MAGE achieves 11.1%–19.3% improvement in worst negative slack (WNS) and 70.0%–74.0% reduction in total negative slack (TNS), significantly outperforming both human experts and Hier-RTLMP, while also enhancing human-like metrics by 6%–48%, demonstrating strong generalization capability.

Abstract

Macro placement still requires substantial manual refinement in industrial physical design flows. We present MAGE (Macro Placement Agentic Engine), a multimodal multi-agent framework for macro placement refinement. MAGE decomposes the macro placement task into a six-phase workflow that combines structured floorplanning rules, visual checks, and iterative refinement. Expert floorplanning knowledge is encoded through natural-language directives and validation criteria, rather than learned from labeled placement data. A tournament-style refinement mode evaluates multiple candidate placements and propagates feedback from higher-quality solutions. We also introduce four metrics for quantifying human-likeness in macro placement: notch score, whitespace score, pocket score, and alignment score. These metrics capture structural properties used by expert designers but not directly measured by conventional PPA metrics. Across nine designs in NanGate45 and GlobalFoundries 12nm enablements, MAGE achieves geometric-mean improvements of 11.1%-19.3% in WNS and 70.0%-74.0% in TNS over commercial macro placers. On the three NanGate45 designs, for which human-expert and Hier-RTLMP baselines are available, MAGE improves WNS and TNS by 18.3% and 72.5% over the human expert, and by 47.0% and 80.4% over Hier-RTLMP, with comparable wirelength and power. On human-likeness metrics, MAGE improves the overall score by 6%-48% over all baselines. Additional case studies on anonymized netlists, unseen designs, dense rectilinear floorplans, and high-utilization settings show that the framework transfers to new placement settings without design-specific retraining.

Problem

• macro placement
• physical design
• human-like layout
• floorplanning
• placement refinement

Innovation

• multi-agent reasoning
• multimodal macro placement
• human-like layout metrics
• language-guided physical design
• zero-shot placement transfer