Curator's Take
AI Commentary
This article marks the first quantum‑hardware demonstration of non‑Abelian adjoint string breaking on a genuinely two‑dimensional SU(2) lattice, using all 156 qubits of IBM’s Boston processor and circuits deep enough (218 two‑qubit layers) to capture real‑time dynamics that match tensor‑network benchmarks. By employing a local Krylov truncation scheme, the authors show that even complex gauge‑field phenomena such as glueball production and string oscillations can be encoded efficiently, suggesting a viable pathway toward simulating strongly coupled quantum field theories beyond one dimension. The result underscores how advances in both algorithmic compression and near‑term hardware scalability are converging to make realistic lattice‑gauge simulations an emerging application for quantum computers.
— Mark Eatherly
Summary
Quantum computers offer the potential to directly probe the dynamics of strongly coupled quantum field theories. As a step towards reaching this potential, local Krylov-based truncations of a pure SU(2) lattice gauge theory on a triangular lattice are constructed. Adjoint strings connected to dynamical gluons are constructed in this truncated theory, and the resonances dominating the long-term dynamics at a large value of the gauge coupling are determined. Local operators are constructed to identify string oscillations and breakings. This is used to perform a quantum simulation of adjoint string breaking on an $8\times8$ and $16\times8$-site lattice with ibm_boston using all 156 qubits. Quantitative agreement with tensor network simulations is obtained for circuits with 7,634 CZ gates with a two-qubit gate depth of 218. In this simulation, the rates of oscillations and glueball production are identified with a distinctly non-Abelian signature of the underlying gauge group.