ML-based Hotspot Selection for OPC / ILT

Hotspot selection determines where expensive ILT should be applied. Machine learning enables fast and accurate hotspot identification.

Input Features

Local pitch and density
2D pattern complexity
Historical CD error
Process window sensitivity

ML Decision Output

Region	Action
Low risk	Model-based OPC only
Medium risk	Extra OPC iterations
High risk	Partial ILT

Advantages

Reduces unnecessary ILT usage
Improves runtime predictability
Maximizes ROI of ILT

Key Takeaway

ML hotspot selection is the gatekeeper for scalable ILT.

↩ Go to OPC / RET Technical Notes Index
↩ Go to Home – Tech Notes Park