The ultimate MOSFET and the limits of miniaturization

After forty years of progress in integrated circuit technology, microelectronics is undergoing a transformation to nanoelectronics. Modern MOSFETs have channel lengths of less than 50 nm, and billion transistor logic chips have arrived. An intensive effort is underway across the world to push MOSFET as short as possible. What are the practical and fundamental limits to MOSFET scaling? To address these questions, we need a sound, conceptual understanding of the nanoscale MOSFET. The conceptual framework that we still use to understand FETs is remarkably similar to the one developed over 40 years ago. In this talk, the author argues that to really understand nanoscale FETs, we should complement traditional FET theory with a new approach. He focuses on the nanoscale MOSFET and describe a very simple but very sound way to understand the device. He examined the practical limits of MOSFETs, whether changing from silicon to something else would change those limits, and whether novel structures, such as nanowire and nanotube MOSFETs, can help. Finally, the fundamental limits of MOSFETs will be identified and the question: "Is there anything better than a silicon MOSFET for digital electronics?" is addressed.