Back to AI News
arxiv

Framework Evaluates AI Agents on Industrial Chip Floorplanning

The paper presents an evaluation framework for AI agents doing chip floorplanning that connects submitted layouts to a prescribed downstream implementation flow, checks geometric validity, measures physical quality and reports resource use. On one industrial design whose 99 macros occupy roughly 71% of a fixed L-shaped region, the authors compare their specialized Nova-SI system with general-purpose engineering agents, obtaining complete physical measurements for 39 layout outputs. Nova-SI reaches a mean of 57.97 on a 0-100 composite quality score without online feedback from the power-grid and placement evaluation, versus 31.17 for the strongest measured general-purpose configuration. Layout-level analysis shows that similar global utilization can conceal substantially different local congestion.

AI Brief

It offers a concrete protocol and empirical evidence for judging AI agents by the physical consequences of their chip layouts rather than by successful tool execution.

arXiv:2611.12048v1 Announce Type: new Abstract: Floorplanning is an early step in turning a logical circuit into a physical chip: it positions large circuit modules and constrains the space available for subsequent logic placement and wiring. These decisions affect congestion and timing, but their physical consequences emerge through downstream implementation. For AI agents, the task requires translating engineering constraints into spatial decisions whose quality cannot be judged by successful tool execution alone. We present an evaluation framework that connects submitted layouts to a prescribed downstream implementation flow, checks geometric validity, measures physical quality, and reports resource use. On one industrial design whose 99 macros occupy approximately 71\% of a fixed L-shaped layout region, we compare our specialized system Nova-SI with general-purpose engineering agents and agents restricted to file access, obtaining complete physical measurements for 39 layout outputs. Under our 0--100 composite quality score, Nova-SI achieves a mean of 57.97 without online feedback from the prescribed power-grid and placement evaluation, compared with 31.17 for the strongest measured general-purpose configuration, GPT-6 Astra. The difference is supported by raw congestion and timing measurements and persists across scoring-sensitivity checks. Layout-level analysis further shows that similar global utilization can conceal substantially different local congestion. Together, the protocol and case study assess agents through the physical consequences of their delivered layouts and provide evidence about system performance on this industrial instance.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!