Choose a country or area to see content specific to your location
W7009E PathWave IC-CAP ANN Modeling Toolkit
A solution for generating a model faster and more efficiently and enhancing the model’s accuracy
Please exit this procurement session to buy.
To buy, contact your procurement team.
The W7009E PathWave IC-CAP ANN Modeling Toolkit substitutes the standard compact model with ANN technology to generate the model quickly and improve the model’s accuracy.
The W7009E PathWave IC-CAP ANN Modeling Toolkit includes:
- Built-in ANN examples in IC-CAP (Diode, Nonlinear resistor, GaAs pHEMT, and Hybrid_Model combining a physical MOSFET model with ANN)
- An application programming interface (API) to use ANN functions in IC-CAP
The Artificial Neural Network (ANN) solution can help modeling engineers improve model accuracy by using an IC-CAP Python script from data input with ANN training and evaluation to export the Verilog-A model. The input for ANN training can be any number of dimensions (e.g., Vg, Vd, T). Instead of traditional physics or equation-based compact models, ANN model is used for extraction. The process is fully automated and faster and more efficient. The output can be either an equation or Verilog-A model, and the convergence can be better and faster than some detailed compact models. The ANN solution can help modeling engineers who need a model for new technologies where they don’t have a compact model yet or the existing models are not accurate enough and need to quickly generate a device model.