What is Optical Computing?
Exploring the Possibilities of Computing with Photons
Introduction
Most of us are aware that the computers we use in everyday life operate using electrons and conductors / semiconductors. Optical computers differ in that they utilize photons of light through fiber optics, holographic components, beam splitters, and optical transistors and logic gates. Since light travels faster than electricity, this greatly increases the speed of optical computers in comparison to their electronic counterparts.
So how exactly do these computers work?
At the core of optical computing lies the optical transistor (also called an optical switch or light valve). A typical optical transistor works by affecting one light source with another within an operating medium with a non-linear refractive index, resulting in an output state of high or low. These can be combined to create logic gates that perform AND, OR, and other standard logic operations.
Data storage is accomplished using holographic memory, which is currently under development, with companies such as Microsoft revealing early prototypes that show great potential. These devices operate by utilizing two sources of light, a reference beam and a data beam, which create an interference pattern within a crystal lattice that can then be read out by cameras and transmitted to other components.
Performance
Optical computing promises incredible performance over conventional computing with electrons. Researchers at NASA theorize that a computation that would take a conventional computer about 11 years to complete would take an optical computer roughly one hour.
Light travels much faster than electrons (299,792,458 m/s versus the average drift velocity of electrons through copper at 100,000 m/s). Additionally, the cycle period of an optical transistor (how fast it can switch on and off) is much shorter than that of a standard transistor — roughly 1/1000th of the time.
Also, whereas conventional computers operate with a single on/off state per wire, optical computers can modulate multiple frequencies of light within one light path, allowing for much more complex operations to be performed.
These factors contribute to the incredible speed and performance offered by optical computing in the future.
Availability
Optical computing is not yet readily available and is generally found within laboratory and other R&D environments, however some companies such as Lightmatter are now offering machines with some optical computing capabilities.
In the future, as optical technologies are perfected in research environments and brought to market, more options will become available, and eventually these technologies will be found in consumer products such as laptop computers and smartphones, offering incredible performance and efficiency over conventional heat-generating electron computing devices.
Applications
Optical computers are especially useful in applications which require very intense computing, such as machine learning and scientific research. In general, basic operations can be performed by electron-based computers with ease, so conventional machines are usually enough to handle day-to-day tasks such as consumer applications like social media and content creation.
That aside, the future of optical computers will likely include consumer devices, opening the possibility of more intelligent devices with stronger capabilities in machine learning and augmented reality for example.
Conclusion
Optical computers, while still relatively new and largely confined to research environments and scientific computing applications, nevertheless promise a bright future for computing in general. As advancements are made in the design and production of these technologies, optical computers will become more readily available in various professional and consumer markets.
