Giant Photonic Processor Solves Graph Issues
• Physics 16, s64
A quantum photonic system can carry out some real-world duties extra effectively than classical computer systems.
Quantum computers can outperform their classical counterparts when solving certain computational problems, but the average off-the-shelf laptop is still far better for most tasks. Now Chao-Yang Lu, Jian-Wei Pan, and their colleagues at the University of Science and Technology of China have used a quantum computer based on a photonic network to solve two graph-theory problems [1]. The consequence extends the listing of duties for which in the present day’s noisy quantum computer systems provide a bonus over classical computer systems.
Beforehand, the group used their photonic processor—a 144-optical-mode interferometer—to unravel an issue referred to as Gaussian boson sampling (GBS); see Viewpoint: Quantum Leap for Quantum Primacy. A GBS resolution is a prediction—or “pattern”—of the likelihood distribution of photons recorded throughout the community’s 144 output detectors when these photons are injected into the community one after the other.
This sampling strategy hyperlinks mathematically to graph issues that mannequin pairwise relationships, outlined by matrices, between objects. The researchers used their photonic processor to implement search algorithms outlined by two such issues. By treating every of the processor’s output ports as a graph vertex and every detected photon as a subgraph vertex, they decided which subgraph mapped to the answer. Their processor arrived at an answer after acquiring 221,891 samples, every of which corresponded to a selected distribution of as much as 80 detected photons. Every pattern would require 700 seconds on the world’s quickest supercomputer utilizing a precise algorithm.
Earlier claims of quantum benefit have been challenged by options that the quantum pc was not competing in opposition to the best-possible classical algorithm for the duty. Whether or not the group’s quantum processor will nonetheless yield a bonus over classical algorithms optimized for fixing graph issues is an open query.
–Rachel Berkowitz
Rachel Berkowitz is a Corresponding Editor for Physics Magazine based mostly in Vancouver, Canada.
References
- Y. H. Deng et al., “Fixing graph issues utilizing Gaussian boson sampling,” Phys. Rev. Lett. 130, 190601 (2023).
