In recent years, the acceleration of computing has attracted much attention due to the rapid development of practical applications such as finance, machine learning and circuit design. These applications are difficult for ordinary personal computers to solve, which requires new technologies to achieve mass computing.  

A research team led by Dr. LIU Hongjun from the Xi'an Institute of Optics and Precision Mechanics (XIOPM) of the Chinese Academy of Sciences (CAS) proposed a novel design to achieve the accuracy-enhanced coherent Ising machine (CIM) by using the quantum adiabatic theorem. Results were published in Optics Express.

Traditional approaches to accelerate computational inevitably have large size, which severely restrict their further applications. While the CIM, based on optical parametric oscillator network, has been recently proven to be the most promising method, because it can take the advantages of high speed and parallelization of the light.

The researchers were inspired by time-multiplexed degenerate optical parametric oscillators (DOPOs) and measurement-feedback approach. DOPOs represent an Ising spin and the in-phase amplitude, which can enhance the performance of CIM.

They calculated the continuous time-dependent Ising Hamiltonian to investigate the adiabatic evolution on networks of DOPOs. By randomly flipping a few spins at the end of each adiabatic step to decrease the Ising energy, the machine can continue evolving towards the Ising ground state.

They also investigated the machine performance depends on the annealing time and results indicate that the time-dependent Hamiltonian during at least 20 evolution steps to ensure the effectiveness of the adiabatic evolution.

To see how the performance of the proposed CIM scales to larger problem size, the researchers solved the sparsely connected G-set instances and fully connected complete graphs with the number of vertices ranging from 800 to 2,000. The outcomes showed the performance improvement of the proposed CIM and computational accuracy of the proposed CIM was also increased.

The proposed CIM can be extended to solve more complex Ising Hamiltonian and physical models in the future.

This work was supported by the National Natural Science Foundation of China.

The schematic of the CIM with measurement-feedback control. (Image by XIOPM)