IEEE Spectrum November 27, 2018
Although many complicated optimization problems can be tackled by optimization, there is still a lot of room for improvement in the methods we use to solve a large fraction of optimization problems. Many common optimization problems, including scheduling and route-finding problems, can be easily converted into Ising optimization problems. Researchers at Stanford University are working to build special-purpose optical machines to solve Ising optimization which is finding the lowest energy state of a collection of spins. Key to their prototype system’s ability to map a spin onto a pulse of light is an optical parametric oscillator (OPO) which produces light that is either exactly in or exactly out of phase with respect to some reference light. According to the researchers their results have been encouraging but have much to learn before they can even say whether this optical approach will ever be able to beat a conventional processor in solving optimization problems of a practical nature…read more.
An optical Ising machines Illustration: Mark Montgomery