Models

The ANT_NanoSOI model is of a 220-nm-thick silicon-on-insulator process using advanced electron-beam lithography.

During the fabrication process, certain physical phenomena such as corner rounding and loss of features below approximately 60 nm can occur. Corner rounding is a common effect in lithography where sharp corners of the design get rounded off during the fabrication process. Similarly, due to the limitations of the fabrication process, features that are smaller than approximately 60 nm may not be accurately reproduced and can be lost.

However, these effects can be partially corrected for with our ANT_NanoSOI corrector model. The model has been trained to anticipate these fabrication artifacts and can adjust the design accordingly to mitigate their impact. This allows for more accurate fabrication outcomes that are closer to the original design.

prediction = pf.predict(
  device=device,
  model_name="ANT_NanoSOI",
  model_tags="v5-d4"
)

prediction = pf.binarize(prediction)  # optional

The ANT_SiN model is of a 400-nm-thick silicon nitride process that employs advanced electron-beam lithography. This model is unique in that it tends to exhibit more pronounced corner rounding and angled sidewalls. These effects can significantly impact the performance of the fabricated devices.

To account for these effects, we have separated this model into two parts: upper and lower. This separation allows us to model and correct for the corner rounding and angled sidewalls more accurately.

Despite these challenges, our ANT_SiN model is designed to anticipate these fabrication artifacts and can adjust the design accordingly to mitigate their impact. This ensures that the final fabricated design is as close as possible to the original, despite the inherent variations in the fabrication process.

prediction = pf.predict(
  device=device,
  model_name="ANT_SiN",
  model_tags=["v5-d0-upper", "v5-d0-lower"]
)

prediction = pf.ternarize(prediction)  # optional