Causal AI For AMS Circuit Design: Interpretable Parameter Effects Analysis

2026-03-24

AI Summary

Analog-mixed signal (AMS) circuits pose significant modeling challenges for conventional data-driven approaches due to their highly nonlinear and continuous nature, resulting in a pronounced gap between design parameters and circuit performance. This work introduces causal artificial intelligence into AMS circuit design for the first time, proposing a framework that constructs directed acyclic graphs (DAGs) via causal discovery algorithms and quantifies parameter influence using average treatment effects (ATE). This enables interpretable parameter ranking and “what-if” predictions. Experimental validation on three operational amplifier designs fabricated in TSMC 65nm CMOS demonstrates that the proposed method achieves ATE prediction with mean absolute error below 25%, substantially outperforming neural networks—which exhibit errors exceeding 80% and frequently mispredict effect signs—thereby overcoming the limitations of black-box models in sign accuracy and error control.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Analog-mixed-signal (AMS) circuits are highly non-linear and operate on continuous real-world signals, making them far more difficult to model with data-driven AI than digital blocks. To close the gap between structured design data (device dimensions, bias voltages, etc.) and real-world performance, we propose a causal-inference framework that first discovers a directed-acyclic graph (DAG) from SPICE simulation data and then quantifies parameter impact through Average Treatment Effect (ATE) estimation. The approach yields human-interpretable rankings of design knobs and explicit 'what-if' predictions, enabling designers to understand trade-offs in sizing and topology. We evaluate the pipeline on three operational-amplifier families (OTA, telescopic, and folded-cascode) implemented in TSMC 65nm and benchmark it against a baseline neural-network (NN) regressor. Across all circuits the causal model reproduces simulation-based ATEs with an average absolute error of less than 25%, whereas the neural network deviates by more than 80% and frequently predicts the wrong sign. These results demonstrate that causal AI provides both higher accuracy and explainability, paving the way for more efficient, trustworthy AMS design automation.

Problem

• Analog-mixed-signal circuits
• Causal AI
• Design parameter interpretation
• Performance modeling
• Explainability

Innovation

• Causal Inference
• Average Treatment Effect (ATE)
• Directed Acyclic Graph (DAG)
• Interpretable AI
• Analog-Mixed-Signal (AMS) Design