Helios-1 quantum computing chip
How many
Researchers at quantum computing company Quantinuum used the new Helios-1 quantum computer to simulate a mathematical model that has long been used to study superconductivity. These simulations are quite accessible to conventional computers, but this advance sets the stage for quantum computers to become useful. materials science tools.
Superconductors conduct electricity with ideal efficiency, but at present they operate only at temperatures too low to be of practical use. For decades, physicists have tried to figure out how to tweak their structure to make them work at their best. room temperatureand many believe the answers can be found in a mathematical system called the Fermi-Hubbard model. This potential makes it one of the most important models in all of condensed matter physics, says a Quantinuum spokesman. Henrik Dreyer.
Conventional computers can perform exceptional simulations of the Fermi-Hubbard model, but have difficulty handling very large samples or cases where the materials it describes change over time. Quantum computers have the potential to eventually achieve more. Now Dreyer and his colleagues have carried out the largest simulation of the Fermi-Hubbard model on a quantum computer to date.
They used Helios-1, which has 98 qubits made of barium ions, each of which is controlled by lasers and electromagnetic fields. To run the simulation, the researchers manipulated the qubits through a sequence of quantum states and then read the output, measuring their properties. Their simulation included 36 particles called fermions, which are exactly the type of particles that exist in real superconductors and are mathematically described by the Fermi-Hubbard model.
For a superconductor to work, fermions must form pairs, and experiments have shown that such pairing can sometimes be initiated by exposing the material to a laser. The Quantinuum team simulated this scenario: They hit the qubits with a laser pulse and then measured the resulting states, looking for evidence of pairing of the simulated particles. The simulations did not exactly replicate the experiments, but they captured a dynamic process that is difficult for traditional computer methods when applied to more than a few particles.
Dreyer says the new experiment is not rigorous proof that Helios 1 has an advantage over all possible traditional computing approaches, but studying classical simulation methods has convinced his team that a quantum computer can compete. “For the methods we tried, it was impossible to reliably get the same results, we were looking at a couple of hours of work on a quantum computer and a big question mark on the classical side of things,” he says. In other words, the team estimated that classical computation times were so long that it was difficult to say when they would be comparable to Helios' work.
Captured ions serve as qubits in the Helios-1 chip
How many
No other quantum computers have yet simulated the pairing of fermions to achieve superconductivity, and the team attributes their success to Helios hardware. David Hayesalso of Quantinuum, says Helios qubits are exceptionally robust and excel at benchmarking tasks common in the quantum computing industry. In preliminary tests it could also withstand experiments with error-tolerant qubits, including connecting 94 of these special qubits through quantum entanglement, a record for any quantum computer. Using such qubits in future simulations could make them more accurate.
Eduardo Ibarra Garcia Padilla Harvey Mudd College in California says the new results are promising but still need to be carefully compared with modern classical computer simulations. He says the Fermi-Hubbard model has been of great interest to physicists since the 1960s, so it's exciting to have a new tool to study it.
Exactly when approaches like those used in Helios 1 will become true competitors to the best conventional computers is anyone's guess, since many details still need to be worked out, he says. Steve White at the University of California, Irvine. For example, he says there are challenges to getting quantum computer simulations to start with the right set of qubit properties. However, White says quantum simulations can be a complement to classical ones, especially for the dynamic or changing behavior of materials.
“They will soon become useful tools for condensed matter modeling. [physics]”, he says. “But they are still in the early stages, and the computing barriers are still ahead.”
Link: arXivDOI: 10.48550/arXiv.2511.02125
Topics:
