FEDERATED LEARNING MODELS FOR PRIVACY-PRESERVING TEST DATA SHARING ACROSS SEMICONDUCTOR DESIGN ECOSYSTEMS
Abstract
The semiconductor design industry relies heavily on collaborative innovation, which often necessitates the sharing of sensitive test data across various stakeholders. However, data privacy, intellectual property protection, and compliance with data regulations remain critical challenges. Federated Learning (FL), a decentralized machine learning paradigm, presents a promising solution by enabling model training across distributed datasets without requiring raw data exchange. This paper explores the application of Federated Learning models for privacy-preserving test data sharing within semiconductor design ecosystems. Our analysis suggests that FL not only strengthens data confidentiality but also fosters collaborative performance enhancements in multi-organization semiconductor workflows.