Asynchronous Processor Design Evolution
This paper discusses how clockless design methodologies can improve the power and silicon efficiency of processors manifold. For fully programmable processors, moving to a clockless design within the same silicon technology can provide power consumption benefits similar to those accruable moving down four geometry nodes like moving from 90nm to 22nm. Additionally, a clockless implementation can also be quite efficient from a silicon perspective. The techniques described herein are technology node independent and the gains achievable are sustainable from one geometry node to another.