View random article

What Is Reversible Computing?

As computer devices grow in speed and capacity, the problem of overheating becomes more and more apparent. The energy dissipated by the device must stay within a particular limit, or it may need a more powerful cooling apparatus—which entails extra cost, and extra space.

To understand the problem, we need to look at the KT Barrier: essentially, the quantity of energy. This is computed by multiplying the temperature of the computing environment (generally room temperature, or ~300 Kelvin) by Boltzmann's constant.

In other words, to get past this barrier, we need to either lower the computer’s temperature or create “thermodynamically reversible computers” that don’t give off that much heat. The latter is far more acceptable a solution. That’s because the temperature of the room, even at its lowest possible levels, don’t significantly make much headway in cooling a very fast, very powerful, but very hot machine. Rather than cooling the room, it is necessary to create a machine that doesn’t create that much heat.

Unfortunately, reversible computing is quite complex. Experts are looking into designing “compilers” which can maximize the efficiency of the machine. In other words, the secret lies in reversible architectures. These are particularly crucial in computers that have very small and fast logic gates, as in the case of nanocomputing.

Efforts to solve this reversible computing challenge are currently being led by MIT. The endeavor has been labeled “The Pendulum Project” which focuses on creating the architectures that would eventually make reversible computing a reality. If they succeed, then both the power and the efficiency of computers can be unstoppable.

Featured in Technology