Irreversibility and heat generation in the computing process
IBM Journal of Research and Development
What can we do with a linear optical logic gate?
Information Sciences: an International Journal
| Hi-index | 0.00 |
The thermodynamic "permission" to build a device that can evaluate a sequence of logic operations that operate at zero energy has existed for about 40 years. That is, physics allows it in principle. Conceptual solutions have been explored ever since then. A great number of important concepts were developed in so doing. Over the last four years, my colleagues and I have explored the possibility of a constructive proof. And we finally succeeded. Somewhat unexpectedly, we found such a proof and found that lossless logic systems could actually be built. And, as we had anticipated, it can only be implemented by optics. That raises a new question: Might an optical zero-energy logic system actually be good enough to displace electronic versions in some cases? In this paper, I do not even try to answer that question, but I do lay out some problems now blocking practical applications and show some promising approaches to solving them. The problems addressed are speed, size, and error rate. The anticipated speed problem simply vanishes, as it was an inference from the implicit assumption that the logic would be electronic. But the other two problems are real and must be addressed if energy-free logic is to have any significant applications. Initial steps in solving the size and error rate are addressed in more detail.