Light signals are not helpful for us to see the world but can be key in solving great computational problems of the day. According to a new study, light signals when multiplied can be instrumental with applications in graph theory, neural networks, artificial intelligence and error-correcting codes. This new proposed computation method can be revolutionary in the field of creating ultra-fast optical computers.
These are computers that use laser or diode produced photons instead of analogue computers which use electrons for computation.
Photon particles are practically mass-less and hence they travel much faster than electron which has a considerate mass. Therefore, the optical computer will be super-fast and also energy efficient. They would be able to process information through multiple temporal or spatial optical channels.
The digital computer speaks and understand binary, i.e. ones and zeroes. The optical computer will understand light signals instead. Two light waves will originate from two different sources and then projecting the result onto '0' or '1' states.
However, using light can cause a nonlinear problem as ‘unknowns’ might arise from while interacting multiplicatively. But the team led by Professor Natalia Berloff think “optical systems can combine light by multiplying the wave functions describing the light waves instead of adding them and may represent a different type of connections between the light waves.”
They used polaritons, which are half-light and half matter, to extend their idea into light pulses in a fibre. These can create tiny pulses or blobs in space which will be overlapped with one another in a nonlinear way.
The team discovered that the key ingredient is how the pulses are coupled with each other. If the coupling and light intensity is right, the light multiplies, affecting the phases of the individual pulses, giving away the answer to the problem. “This makes it possible to use light to solve nonlinear problems,” said Nikita Stroev, associated with the study.
Wave functions multiplication will determine the phase of light signal. They observed there will be no need to project continuous light for “0” and “1” variables. “Instead, the system tends to bring about these states at the end of its search for the minimum energy configuration,” Stroev added. This is achieved by multiplying the light signals.
Berloff suggested they can start identifying problems which can be directly detected by physical processors. Of course, many challenges lie ahead of them. But they assure their model has superior qualities than electronic computers with, “noise reduction, error correction, improved scalability, guiding the system to the true best solution are among them.”