Enhancing VLSI Design and Manufacturing using AI Techniques in MOSFET

Making and designing MOSFET transistors the traditional way has some serious issues. Engineers currently have to do a lot of manual labor and use crude estimates, and that usually results in errors in forecasting how well the transistors will function. The proposed work applied intelligent computer algorithms, namely Deep Neural Networks and Linear Regression, to learn automatically the optimal transistor designs. The suggested system will predict threshold voltages 15% more precisely by approximating drain current with 22% lower errors and also measure subthreshold swing 18% better. This system even developed a simple-to-use interface so that users are able to clean up their data, train the models, and make predictions in real-time without having to be AI experts. This reduces the number of design attempts by 40% and achieves the right design the first time 35% more frequently. Basically, this system showing how artificial intelligence can completely change the game for chip design - making it faster, cheaper, and makes less dependent on human guesswork.