Traditional computers struggle with NP-complete problems, which grow exponentially in complexity. According to a study published in Advanced Photonics, a group of researchers from Shanghai Jiao Tong University has made progress in this field by creating a reconfigurable three-dimensional integrated photonic processor specifically intended to address the subset sum problem (SSP), a well-known NP-complete issue.
As technology develops, the limitations of traditional electronic computers become more noticeable, particularly when dealing with challenging computing problems. Some of the hardest problems in computer science are NP-complete issues.
These problems have important ramifications for industries such as manufacturing, transportation, and biomedicine. Researchers are looking for alternatives to conventional computing techniques to find more efficient solutions, and optical computing looks promising.
The researchers built a photonic chip of 1,449 standardized optical components using a novel femtosecond laser direct writing technology. This technology enables fast prototyping and increased design flexibility, essential for handling the SSP's complexity.
The SSP explores whether a particular subset of numbers can add up to a specified target. The researchers can compute by encoding the behavior of light on their photonic processor. The processor works by enabling a light beam's photons to explore all possible paths, offering answers in parallel simultaneously.
The processor's ability to solve several instances of the SSP with 100 % dependability demonstrates that this design maintains great precision and speeds up computation.
The SSP is not the only possible use for this technology. The possibility of adapting the processor's reconfigurable nature to tasks like optical neural networks and photonic quantum computing suggests a versatile future for photonic systems. Surprisingly, this new processor has outperformed its electrical competitors, especially regarding computation speed and efficiency as problem size grows.
This advancement opens the door to solving computationally challenging issues on a larger scale and represents a major step in utilizing light's capabilities for practical computing. As scientists continue investigating optical computing, this discovery may change how we tackle difficult problems in various scientific and industrial fields.
Journal Reference:
Xu, X.-Y., et al. (2024) Reconfigurable integrated photonic processor for NP-complete problems. Advanced Photonics. doi.org/10.1117/1.ap.6.5.056011.