A team of engineers has developed a new sort of integrated computer circuit based on highly strange interactions between light waves and sound waves. The technology, which is described in the current Nature Communications, may have big implications for quantum computing.
Sound is a much dumber, slower form of wave than light. Photon particles, being the lightest and fastest thing in the whole of existence, might not normally even deign to interact with sound in any super-meaningful way. They're not even really in the same ballpark as phenomena—sound is macro-scale, consisting of periodic movements within large collections of different sorts of materials, while light is just light. Light waves are periodic movements within light itself.
The new chip, which comes courtesy of a team led by University of Minnesota computer engineer Mo Li, generates and traps sound and light waves together. This confinement is what allows the acoustic waves to efficiently control the optical waves; smashed together into a superthin layer, acoustic wavelengths can actually become smaller than optical wavelengths.
To put that into perspective, consider that a middle-C musical note has a frequency of 260 Hz, while the frequency of blue light would be about 650 THz, or 650,000,000,000,000 hertz.
Within this thin layer, acoustic waves are boosted to over 10 GHz, which is almost the frequency of microwaves. That's pretty nuts.
The general idea behind acousto-optic devices dates back to the 1920s and, as the current paper notes, there now exists entire families of components based on the phenomenon: modulators, frequency shifters, beam deflectors and scanners, tunable filters, and so on. These devices tend to be rather bulky compared to the optical devices they'd ideally be integrated with, and they've also only been able to achieve frequencies in the megahertz range.
That's far too sluggish for the ideal optical world, but acousto-optics have hardly remained a lost cause. The slowness of acoustic waves has been a function of manufacturing limitations, not of fundamental physics/engineering. Those limitations have been officially beaten.
"With the significant advances in nanofabrication, inter-digital acoustic transducers can now be readily fabricated with sub-micron linewidth to generate surface acoustic waves with ultrahigh frequency up to tens of GHz," the paper notes. "At the same time, nanophotonic waveguides and cavities with very high quality factors have been developed to confine light in sub-wavelength scale with extremely high optical power density." The technology has arrived.
Acoustic waves within an integrated chip mean something a bit different than the acoustic waves that normally collide with our ears. They don't propagate through the air, but along the surface of a silicon substrate, like earthquake waves radiating outward from an epicenter.
The basic idea of an optical integrated circuit is that light, lasers operating often within the visible spectrum, take over the duties usually reserved for electricity. The result should be be smaller and faster chips, ditching the inherent size limitations of dealing with electrons. Sound waves are able to act as a sort of pliable, generic medium with which it's possible to control light waves in all sorts of different ways: frequency modulation, deflection, signal processing, etc. Instead of wires, we just have these beams being controlled and filtered by sound waves.
At this point in computer evolution, we don't really have many other options for making things smaller and faster. Moore's Law itself may depend on this very technology.
A pre-publication, open-access version of Li's paper can be found at arXiv.