Performance comparison of artificial intelligence networks in nanoscale MOSFET modeling

In this paper, different types of artificial intelligence networks were compared in order to simulate the nonlinear behavior of nanoscale MOSFETs. The accuracy of the approaches in determining the device drain current and the training time were discussed. The training data was generated in Hspice environment and imported in Matlab7.5 for simulation. Finally, optimized structures for accurate and fast device simulations were introduced individually.