Semiconductor Devices: Moore Marches On

he tremendous growth in semiconductor technology has been based on the well-behaved Si−SiO2 system. In 1965, Gordon Moore, cofounder of Intel and arguably APL’s most famous alumnus, saw the future. His prediction, now popularly known as Moore’s Law, states that the number of transistors on a chip doubles approximately every 2 years. Up until now this “law” has been followed by evolutionary progress of the basic technology. This progress has reached its limit. Major material modifications to the gate and drain regions of active devices will allow improvements in Si-based components to continue until 2020, but completely new concepts and technologies will be required to allow Moore’s Law to hold after that. These technologies include graphene and carbon nanotube-based devices, single-electron transistors, spintronics, and quantum computing. None of these are at a technology readiness level to take over, although spintronics has a good start because it already forms the basis of multi-billion dollar industries in magnetic read heads and magnetic random access memories.