Researchers have developed a novel physics-informed graph attention network (GAT) surrogate that operates directly on TCAD meshes to simulate FinFET device behavior. This approach predicts electrostatic potential and carrier quasi-Fermi levels at each mesh node, embedding carrier-transport physics into the training objective. The surrogate demonstrates size generalization, allowing models trained on smaller meshes to be applied to larger, more complex device geometries. Benchmarked against Sentaurus Device, the GAT surrogate achieves orders of magnitude higher throughput while maintaining high accuracy, enabling efficient design space exploration for multi-fin tri-gate FinFETs. AI
IMPACT Enables faster and more efficient design space exploration for advanced semiconductor devices.
RANK_REASON The cluster contains an academic paper detailing a new machine learning model for scientific simulation. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
