Researchers at Nikhef Maastricht have used a quantum computer to identify particle tracks in data from the LHCb detector.
The results appear to be comparable to those obtained using conventional reconstruction methods, they write in a new article in Nature’s *Communication Physics*.
PhD candidate Xenofon Chiotopoulos and colleagues describe a successful experiment involving track reconstruction on two different quantum computers. Their conclusion is that the algorithm works for simplisme detector image.
Quantum computers operate on the basis of qubits, which are not just 1 or 0, but a superposition of both. In principle, this allows countless versions of a computational task to be executed simultaneously. Quantum computations require algorithms that are very different from those used in digital computations.
In a particle accelerator experiment such as LHCb at CERN, protons collide, after which the newly released particles fly through the detector and trigger sensors there.
The task of track reconstruction is to correlate the sensor signals from the detector so that they reveal traces of collisions that can be physically investigated.
Classical track reconstruction works step by step, starting with an initial signal near the collision and moving toward plausible related signals further down the detector. The goal of quantum computing is to compute that entire pattern in a single operation across countless possible relationships and then filter the best answers from the result.
LHCb Maastricht had previously worked on quantum algorithms for this type of problem. The researchers have now mathematically simplified the problem of tracks in the LHCb into a system of linear equations. They are using the existing HHL quantum algorithm for solving linear equations to find tracks.
To do this, they analysed image from simplyfied collisions using two quantum computers—one on an online quantum platform at IBM and one at Quantinuum. The analysis involved processing signals from five detector layers.
The results of the experimental calculations are promising, says Chiopoulos. Both tests show that the designed algorithm recognises traces in a simplified application involving a few particles across five layers of sensors. A classical algorithm finds the same traces in many more computational steps, according to the researchers.
Quantum computing is an option for the distant future, says Chiopoulos. “Now is the time to explore these techniques.”
