The road to Exascale will likely require processor power efficiency to increase by 1000x. To get there, researchers at the University of South Florida are exploring a revolutionary approach called reversible computing based on the original research by Charles H. Bennett of IBM.
In a reversible computing model, the computational process is time-invertible. In other words, if a reversible computation produces a series of outputs, then the inverse computation on these outputs should reproduce the original states.
How would this work? According to this Wikipedia post, “we would need to precisely track the state of the active energy that is involved in carrying out computational operations within the machine, and design the machine in such a way that the majority of this energy is recovered in an organized form that can be reused for subsequent operations, rather than being permitted to dissipate into the form of heat.”
When conventional logic gates produce several outputs, some of these are not used and the energy required to generate them is simply lost. These are known as garbage states. “Minimization of the garbage outputs is one of the major goals in reversible logic design and synthesis,” say Himanshu Thapliyal and Nagarajan Ranganathan at the University of South Florida.
At this time, no such reversible logic gate exists, so this work is entirely theoretical. However, Thapliyal and Ranganathan are looking at how they might apply this approach using the emerging technology of quantum cellular automata. In such a system, no bits are erased and therefore no energy is dissipated.