1: \begin{abstract}

2: With the feature size continuously shrinking in advanced technology nodes,

3: mask optimization is increasingly crucial in the conventional design flow,

4: accompanied by an explosive growth in prohibitive computational

5: overhead in optical proximity correction (OPC) methods.

6: Recently, inverse lithography technique (ILT) has drawn significant attention

7: and is becoming prevalent in emerging OPC solutions.

8: However, ILT methods are either time-consuming or in weak performance of mask printability and manufacturability.

9: In this paper, we present DevelSet, a GPU and deep neural network (DNN) accelerated level set OPC framework for metal layer.

10: We first improve the conventional level set-based ILT algorithm by introducing the curvature term to reduce mask complexity and applying GPU acceleration to overcome computational bottlenecks.

11: To further enhance printability and fast iterative convergence,

12: we propose a novel deep neural network delicately designed with level set intrinsic principles to facilitate the joint optimization of DNN and GPU accelerated level set optimizer.

13: Experimental results show that DevelSet framework surpasses the state-of-the-art methods in printability and boost the runtime performance achieving instant level (around 1 second).

14: \end{abstract}

15: