THE QUANTUM INTELLIGENCER

A curious assessment crosses my desk this morning, one that asks how far a quantum state may be preserved while doing nothing at all. The authors of an arXiv preprint have compared the Rigetti Cepheus-1-108Q and the IBM Heron-r2 processors, using a deterministic state-transfer protocol that routes a prepared state across the physical lattice and measures fidelity against a classical limit. In the highlighted IBM case, they report an isotropic radius of ten and a successful path of swap distance twenty-seven; the Rigetti machine, by contrast, shows an isotropic radius of but one, though selected routes reach a swap distance of eight. The distinction between uniform spatial reliability and best-route performance is sharp, yet the authors properly caution that these quantities are empirical, conditional on the evaluated route families, the finite-shot decision rule, and the calibration state; they are not architecture-wide constants. Such a result, while narrow, is worth cataloguing for the engineering record. Let the careful reader take this assessment for what it is: a narrow measurement, not a verdict. The figures establish nothing about which machine might serve a practical computation, for the protocol examined is a single-qubit state transfer, a deliberately elementary operation. It says nothing of entangling gates, error correction, or the many benchmarks by which a quantum processor earns its reputation. Nor do these numbers speak to error rates or coherence times; they are particular to the route families evaluated, the finite-shot decision rule, and the calibration state on the day of execution. That the IBM instrument achieved an isotropic radius of ten while the Rigetti reached but one, though it sustained longer paths, is a contrast in the shape of reliability under these specific conditions, not a ranking of the two companies' lines. One might also note that the protocol is not truly doing nothing: it actively routes the state across the lattice with SWAP gates, so it measures preservation under transport, not idle decoherence. As a preprint, it has not undergone peer review, and its numbers must be received provisionally. What it does establish is a useful method for probing spatial reliability on a processor, and a reminder that such measurements are always conditional on the instrument's state, never fixed constants of the architecture. —Ada H. Pemberley Dispatch from The Prepared E0