Google LLC today detailed two internal research projects focused on using quantum processors to study quantum materials, a category of materials that can’t be explained by classical physics.

In the first project, the company’s researchers found a way of simulating quantum materials with more accuracy than was possible until now. As part of the other research initiative, they developed a new method for carrying out simulations that could have applications in future experiments.

The term quantum materials refers to a broad category of microscopic objects ranging from sheets of graphene to so-called ultracold atoms, or atoms that are cooled to temperature near absolute zeros. What these objects have in common is that they’re so tiny their behavior is governed by the rules of quantum mechanics, which apply at the atomic and subatomic scales, rather than by the rules of classical physics, which affect large objects.

Researchers study quantum materials using simulations. It’s believed that it would be possible to improve the quality of quantum materials simulations, and thus help researchers make new discoveries, by running them on quantum processors like the ones Google is developing.

Currently, however, there are several technical challenges standing in the way. One of the main challenges is that it’s difficult to create accurate simulations on quantum processors because the processors are prone to computing errors. The errors decrease the accuracy of the simulations, which in turn limits the ability of researchers to study quantum materials.

In the first research project that Google detailed today, the search giant’s scientists found a method of filtering computing errors and thereby increasing simulation accuracy. They tested the method by using one of Google’s internally developed quantum computers (pictured) to run a simulation of a tiny quantum material wire.

The scientists determined that, when the results of the simulation are visualized and turned into a graph resembling a sophisticated bar chart, it’s possible to distinguish computing errors clearly from accurate data. The process involves encoding the errors into the height of each bar in the bar chart using a mathematical operation known as a  Fourier transform.

Because the height of the bars is fairly simple to measure, researchers can spot processing mistakes with relative ease and filter them so only the accurate simulation data remains. They can then study this data to glean new insights about quantum materials.

Google’s scientists said that they managed to improve simulation accuracy significantly using the method. That’s despite the fact that the simulation in which they tested the method, which represented a tiny quantum material wire, was fairly complicated. “Despite being an 18-qubit algorithm consisting of over 1,400 logical operations, a significant computational task for near-term devices, we are able to achieve a total error as low as 1%,” detailed Charles Neill and Zhang Jiang, senior research scientists at Google’s Quantum AI unit.

In the second project Google detailed today, the search giant developed a new way of studying electrons as part of quantum material simulations. The company’s scientists configured a quantum processor’s qubits to “act” as electrons in order to simulate their physical properties. By carrying out a certain sequence of computing operations, the qubits could simulate changes in the behavior of the electrons they represented.

“Our results provide an intuitive picture of interacting electrons and serve as a benchmark for simulating quantum materials with superconducting qubits,” Google’s scientists wrote.

The two projects are also significant because current quantum processors can perform only a narrow set of tasks. By finding ways of applying the processors to more tasks, in this case simulation, researchers can get closer to developing general-purpose quantum computers capable of running many different software applications. 

Photo: Google