A team led by Eleanor Crane at MIT developed a digital toolkit for simulating light and matter interactions using quantum hardware. Published in Physical Review X, the study uses microwave cavities to represent bosons alongside qubits for fermions. This hybrid approach helps computers work better without needing extensive error correction.
Many of nature's most important processes depend on how particles of matter interact with light and other force-carrying particles. So far, however, it has proven difficult for conventional computers to determine how individual particles behave in these interactions.
Now, a team led by Eleanor Crane at MIT has developed a digital toolkit for simulating these interactions using quantum computing hardware, which could allow such machines to outperform their classical counterparts without the need for extensive error correction.
Their research has been published in Physical Review X.
Differing simulation demands
In physics, particles are broadly divided into two families: fermions such as electrons and quarks, which make up matter, and bosons such as photons and gluons, which carry the forces between them.
In certain cases, fermions and bosons interact particularly strongly, such as the quarks and gluons within protons and neutrons or the electrons that pair up via vibrations in the atomic lattices of superconductors. Here, the computing power a conventional computer needs to track each particle's behavior grows exponentially with the size of the system.
To an extent, quantum computers offer a solution to this problem. When observed, their building blocks, called "qubits," can exist in one of two possible quantum states. Since no two fermions can occupy the same quantum state, each state is either empty or filled, making qubits a natural platform for simulating them.
Bosons, however, can crowd into the same quantum state in any number. In a simulation, this number must be capped and spread across several qubits, demanding long, error-prone sequences of operations.
Integrating cavities
In their study, Crane's team tackled this problem using hardware that already contains bosons. In superconducting quantum computers, qubits are often linked to microwave cavities: small metal boxes in which an electromagnetic field oscillates back and forth. Until now, these cavities have largely played supporting roles, helping qubits to interact with each other and to be read out. In this case, they took a more direct role.
The researchers translated models of interacting particles and fields into operations carried out on both qubits, representing fermions, and cavities, representing bosons. Crucially, the number of steps needed to control a cavity stays roughly the same no matter how many bosons it holds, and every operation is already available on existing hardware.
Fewer steps, fewer errors
To test their approach, Crane's team calculated how such a hybrid processor would simulate a simplified model sharing key features with the theory describing the Higgs boson, while accounting for realistic sources of error.
They found that hybrid hardware making one error in every 1,000 operations could match the accuracy of a qubit-only machine making just one error in every 1 million, a level far beyond today's technology. On this basis, the team proposed an experiment that could outperform both classical simulations and near-term qubit-only machines.
The team's toolkit could help researchers study how electrons interact with vibrations inside superconductors and semiconductors, how light influences chemical reactions, and how particles scatter and evolve in particle physics experiments. It could also be adapted to other platforms, including trapped ions and neutral atoms.
As the hardware it requires already largely exists, their next step will be to put their proposal to the test, potentially opening a new route beyond qubit-only quantum computing.