Co-modeling post-lithography critical dimensions and post-etch critical dimensions with multi-task neural networks
A method is presented for constructing a deep neural network based model to concurrently simulate post-lithography critical dimensions (CDs) and post-etch critical dimensions (CDs) and to improve the modeling accuracy of each process respectively. The method includes generating lithographic aerial images of physical design layout patterns, constructing a multi-task neural network including two output channels, training the multi-task neural network with the training data of the lithographic aerial images, and outputting simulated critical dimension values pertaining to lithography and etch processes.
1. A method for concurrently modeling post-lithography critical dimensions (CDs) and post-etch critical dimensions (CDs), the method comprising:
generating lithographic aerial images of selected physical design layout patterns, the lithographic aerial images employed as training data;
constructing a first portion of a neural network to be shared by two output channels;
constructing a second portion of the neural network to be customized for each of the two output channels;
training the neural network with the training data; and
outputting simulated critical dimension values for lithography and etch processes.
2. The method of claim 1 , wherein the first output channel is a post-lithography CD channel and the second channel is a post-etch CD channel.
3. The method of claim 1 , wherein the neural network is a multi-task neural network.
4. The method of claim 1 , further comprising analyzing correlational relationships between the post-lithography CDs and the post-etch CDs.
5. The method of claim 1 , further comprising applying the neural network to concurrent full-chip modeling of the post-lithography CDs and the post-etch CDs.
6. The method of claim 1 , further comprising fine-tuning hyperparameters of the neural network based on the simulated critical dimension values for lithography and etch processes.
7. The method of claim 1 , further comprising displaying a visual representation of the simulated critical dimension values on a user interface.
8. A method for constructing a deep neural network based model to concurrently simulate post-lithography critical dimensions (CDs) and post-etch critical dimensions (CDs), the method comprising:
generating lithographic aerial images of physical design layout patterns;
constructing a multi-task neural network including two output channels;
training the multi-task neural network with the training data of the lithographic aerial images; and
outputting simulated critical dimension values pertaining to lithography and etch processes.
9. The method of claim 8 , wherein the first output channel is a post-lithography CD channel and the second channel is a post-etch CD channel.
10. The method of claim 8 , further comprising analyzing correlational relationships between the post-lithography CDs and the post-etch CDs.
11. The method of claim 8 , further comprising applying the multi-task neural network to concurrent full-chip modeling of the post-lithography CDs and the post-etch CDs.
12. The method of claim 8 , further comprising fine-tuning hyperparameters of the multi-task neural network based on the simulated critical dimension values for lithography and etch processes.
13. The method of claim 8 , wherein the post-lithography CDs and the post-etch CDs are concurrently modeled for full-chip level evaluation.
14. The method of claim 8 , further comprising displaying a visual representation of the simulated critical dimension values on a user interface.
15. A system configured to train a machine learning based model, the system comprising:
one or more computer subsystems; and
one or more components executed by the one or more computer subsystems, wherein the one or more components include the machine learning based model configured for performing one or more simulations by:
generating lithographic aerial images of selected physical design layout patterns, the lithographic aerial images employed as training data;
constructing a first portion of a neural network to be shared by two output channels;
constructing a second portion of the neural network to be customized for each of the two output channels;
training the neural network with the training data; and
outputting simulated critical dimension values for lithography and etch processes.
16. The system of claim 15 , wherein the first output channel is a post-lithography CD channel and the second channel is a post-etch CD channel.
17. The system of claim 15 , wherein the neural network is a multi-task neural network.
18. The system of claim 15 , wherein correlational relationships between the post-lithography CDs and the post-etch CDs are analyzed.
19. The system of claim 15 , wherein the neural network is applied to concurrent full-chip modeling of the post-lithography CDs and the post-etch CDs.
20. The system of claim 15 , wherein hyperparameters of the neural network are fine-tuned based on the simulated critical dimension values for lithography and etch processes.