Erasure Conversion and the Error Budget: What Detectable Errors Buy a Surface Code
An error whose location is known is far cheaper to correct than an error that must first be found: pure erasure channels tolerate error rates several times higher than Pauli channels, a fact known since the earliest threshold studies. Erasure conversion, engineering a qubit so that its dominant failures announce themselves, therefore buys fault tolerance with information rather than with distance. This paper budgets the purchase. If ninety per cent of a platform's errors are converted to located erasures, the unlocated residue falls from 1.00e-3 to 1.00e-4, and the effective threshold for the mixed channel rises by a factor of about 4 over the purely unlocated figure. Priced through the standard scaling ansatz, the conversion is worth several units of code distance at fixed logical quality, which at algorithmic distances is worth a third or more of the machine. The paper reviews the hardware routes that deliver conversion, dual-rail superconducting qubits, metastable-state ions and loss-detecting atom encodings, and charges each route its conversion overhead.
Authors
- William Alexander Aldred (ORCID: https://orcid.org/0009-0000-1215-577X)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-15
- DOI
- https://doi.org/10.5281/zenodo.24245368
- Primary Topic
- Quantum Computing Algorithms and Architecture
- Type
- article
- Field-Weighted Citation Impact
- 0.00