The Silicon Transistor: Personal Observations

25 Is the decades-long era of exponentially compounding improvements in the cost and performance of computing devices about to end? Can we sustain progress in information technology hardware by developing some new switching device that can be made smaller, faster, and cheaper than the silicon transistor? These were some of the questions raised by a panel session, “Looking Beyond Moore’s Law, A Technical Perspective,” which I had the pleasure to host during a recent CIA/DARPA conference on the Global Computer Industry Beyond Moore’s Law. The conference organizers asked each panelist to discuss a particular view of the future of computing. (See the “Session Panelists” sidebar for a list of panelists and their topic areas.) Following the panel’s theme, the participants summarized the current ideas and prospects for further progress in their respective areas, including molecular devices and quantum computing. Implicit in the choice of panel session titles, speakers, and topics was the idea that progress in the miniaturization of silicon transistors must soon end. However, this view has been prevalent and consistently wrong during my entire technical career. The silicon transistor has remained the forerunner in mainstream computing. Yet, the future of computing is not set in silicon or any other technology. The panelists’ discussion sparked some of my own observations on where I feel the field is going.