AnalogLayoutWorld benchmark tests AI for analog circuit design tasks
AnalogLayoutWorld: An Open Benchmark for Agentic Analog Layout Design
Artificial Intelligence
Summary
Designing analog circuits is hard because it needs specialized knowledge and careful planning. The authors created AnalogLayoutWorld, a set of 40 challenging tasks that test if AI agents can generate layouts of analog circuits. They also provide tools to check if the designs work well physically and electrically. Their experiments show many AI attempts create layouts, but only a few meet strict quality standards. They suggest better reasoning and using feedback effectively help AI handle more complex designs.
What this means in practice
- •For analog circuit engineers: Use the benchmark to test and improve AI tools that automate analog circuit layout design from transistor netlists.
- •For electronic design automation developers: Integrate the layout generation benchmark into verification pipelines to evaluate AI agent effectiveness on complex analog design tasks.
Authors
NovaSilicon Team
Abstract
Coding agents have demonstrated strong capabilities in software development, yet their ability to tackle long-horizon engineering tasks governed by physical constraints remains less understood. We study this question through analog circuit layout, which demands domain knowledge, geometric reasoning, and iterative interaction with design and verification tools. We introduce \textbf{AnalogLayoutWorld}, an open-source agent benchmark with 40 expert-reviewed tasks for end-to-end layout generation from fixed transistor-level netlists. It combines an integrated design environment with independent physical and post-layout electrical verification, as well as objective layout quality metrics co-designed with analog layout experts. Across 13 model--harness configurations, 91.1\% of attempts produce GDS layouts, but only 23.0\% satisfy the benchmark's acceptance criteria. We also evaluate Novara, an internally developed agent system, which achieves the highest reported aggregate score. Our analysis links performance differences to test-time scaling under fixed budgets, suggesting that reasoning efficiency and effective use of verification feedback help sustain design quality as task complexity grows.