A team of scientists at the University of Oxford has set up a new computer storage unit that can be accessed or written by both electrical and optical signals, an unexpected breakthrough that means the feasibility of chip-grade photonology is expected to rise rapidly. Replacing electrons with light is a distinctly more ideal form of signal, which guarantees greater bandwidth and higher power efficiency. However, the complexity of using this fine form of energy makes it difficult to break through research and development.
The only widespread use of photonics is in fiber optic cables, from home Ethernet to optical cables that span the distance between continents, and we can see optical communication.
The use of plate and chip-level steamer makes photonics a hot topic again. In short, optical communication has a lower power threshold, so signals can be sent faster and earlier, with on-chip delays that can be several orders of magnitude lower than electrical signals. The problem is that converting optical signals to electrical signals requires a lot of power and space, offsetting all gains.
In the past, we’ve seen interesting attempts to solve this problem, but new research has solved it in a completely different way. If memory (whether primary, memory, or cache) can accept and output information in both formats, there is no need to convert.
The scientists’ latest lym joint is a non-volatile niobium-based compound at the intersection of the gold electrode and the silicon nitride channel. Electrons flow through gold, and light waves pass through the channel funnel. When either cell hits the unit, the cell can switch between binary or multi-level states.
While technicians are expected to implement the technology for specific purposes in the coming decades, scientists have foreseen how it will help solve current problems, such as new technologies that can promote the development of photon transistors and serve as caches and interfaces for reconfigurable photon circuits and photon neural networks.