Although a large portion of the articles in Quantum Computing Report by GQI cover gate-based quantum computers, we should remind our readers that these are not the only game in town. In particular, there are also many other types of physics-based processors that have a potential to outperform classical computers and gate-based computers for problems the use combinatorial optimization. They include quantum annealers, simulated bifurcation machines, Lightsolver, and the Coherent Ising Machine. GQI believes that combinatorial optimization problems may comprise as much as one-third of the total market for quantum applications. So although these optimization focused machines may not be as general as the gate-based systems, there still is a significant market to make development of these special purpose processors potentially worthwhile.
In the past, we have reported on the research to develop Coherent Ising Machines (CIM) being performed at NTT Research with several other Japanese and North American academic partners. (See our previous articles here and here.) This device operates by sending photon pulses through an optical loop and then using controlled feedback between an outlet coupler and a injection coupler to provide constructive or destructive interference to amplify the pulses with the correct solution and attenuate the others. This CIM architecture provides for all-to-all connections of the photon pulses and it can potentially be extended to hundreds of thousands or even millions of photons. Early work on this concept was performed at Stanford University.
The previous implementation of NTT’s CIM utilized 1 km of fiber optic cable as shown in the picture below.
Although this implementation worked, like many early generation devices it was big, expensive, slow, and not very stable. In a device like this, temperature changes in the room can affect the performance and reliability. So the important development that NTT Research is pursuing now is to implement the CIM device with a chip level implementation using a Thin Film Lithium Niobate (TFLN) technology that the company has been developing. In addition to the loop itself, other components of the CIM can be integrated onto the same chip. In much the same manner as when classical computers improved significantly when the technology went from vacuum tubes to integrated circuits, similar things will happen with with Coherent Ising Machines when they start to us TFLN chips. The resulting system will be much smaller, faster, reliable, cheaper, and easier to build because it will be largely based on semiconductor manufacturing technology.
A conceptual diagram of how the photon loop would be implemented on a TFLN chip is shown below.
Although bulk Lithium Niobate (LiNb03) has been around for a long time and is used in a variety telecom applications, thin film is a relative newcomer and has great promise in many different photonic applications due to its wide optical transparency window, low optical losses, appreciable nonlinear coefficients, and amenability to nanofabrication processes. Historically, silicon on insulator (SOI) or silicon nitride have been used for photonic based quantum computers, but the superior characteristics of TFLN may ultimately change that. NTT is also researching the use of TFLN for photonic neural network processors optimized for machine learning applications.
Currently, much of the supply of TFLN wafers comes from a company in in Jinan, Shandong province, China called NANOLN. So having additional sources of this material in other countries would be helpful for a robust supply chain.
For additional reading about NTT research in the Coherent Ising Machine and also TFLN development, we can refer you to the following sources:
- Quantum Nature of the CIM
- CIM on Chip Demonstration
- Nonlinear Nanophotonics: Towards Few-Photons Interactions
- Statistical Mechanics of High Dimensional Optimization Landscapes in the CIM
- A Fast, Scalable, and Reconfigurable Simulation Platform for the Coherent Ising Machine
- Coherent Nonlinear Dynamics and Combinatorial Optimization
- Neuromorphic in Silico Simulator For the Coherent Ising Machine
- Coherent SAT Solvers: A Tutorial
- NTT R&D Forum 2023 Special session3:Lithium Niobate Photonics In The Era of AI
- Mid-infrared Nonlinear Optics in Thin-film Lithium Niobate on Sapphire
April 17, 2024