Revision 1

Accepted · first revision · record

Title: Has an analog quench simulation exceeded classical computation?
Plain: A 2025 experiment claimed a quantum annealer simulated magnetic dynamics that no supercomputer could match. Classical simulation groups replied with algorithms that may do the same job.
Precise: Transverse-field Ising quenches on the square-lattice, dimerized-cubic, diamond-lattice and biclique spin-glass instances of King et al., Science 388, 199 (2025). Determine whether a classical algorithm reproduces the reported correlation functions to the hardware's accuracy with comparable time and energy. Answer: yes or no for each graph family studied.
Settled by: A side-by-side benchmark of classical methods and the device on the same instances with agreed accuracy targets.
Kind: well-posed
Answer type: yes-no
References: (empty)
Literature status: contested
Status note: Belief-propagation tensor networks (Tindall et al., Science 2026) and time-dependent variational Monte Carlo (Mauron and Carleo, arXiv:2503.08247) disputed the claim, and a 2026 comment argued neural quantum states are competitive on some instances.
Posed since: 2025