A team of physicists in China claimed it has demonstrated a “quantum advantage,” introducing a light-driven quantum computer capable of performing a computation that would take a classic supercomputer an estimated 600 million years to resolve. The team’s photonic quantum computer ultimately took several minutes to complete the boson sampling problem — a computation originally devised in 2011. The result follows Google’s 2019 announcement that its device, dubbed Sycamore, was the first to achieve a clear quantum advantage. Where chip-based superconducting circuits powered Sycamore, though, the Hefei, China-based team relied on photons delivered by a pump laser. To resolve the boson sampling problem, for which solutions increase as influential variables increase, a computer must calculate probability of dispersed boson location. The quantum waves of bosons cause a natural interference that randomly positions the particles. An equation-based method can determine the positioning of bosons. Photons are compositional of bosons, and the team, led by Jian Wei-Pan and Chao-Yang Lu from the University of Science and Technology of China, used photons as qubits in its system. Taking the photons’ spatial positioning and polarization at specific states into account and by merging the two measurements to cause photon interference, the researchers showed the ability to distribute photons in their device. High-efficiency detectors successfully sampled the output distribution represented by the photons, measuring their distribution and helping definitively solve the equation. The photonic quantum computer generated as many as 76 distinct output photon clicks in the process, exhibiting an output state-space dimension of 1030 and a sampling rate that is approximately 1014 faster than using the state-of-the-art supercomputers. In 200 seconds, the device, in other words, reportedly performed a task that would take more than 2.5 billion years for a supercomputer to complete. The researchers named their device “Jiuzhang.” The research was published in Science (https://science.sciencemag.org/content/early/2020/12/02/science.abe8770).
