DEV Community

Alfio Musumeci
Alfio Musumeci

Posted on Edited on Originally published at alfiomus.blogspot.com

Rare-Earth Quantum Memories, Week of Nov 2: No New Record, a Better Scoreboard

#quantum #quantumcomputing #science

Week of November 2, 2026. A consolidation week: the record-holding group published the field's scoreboard, and erbium got new operating-point math. Canonical version with full sources: [link at bottom].

The europium memory setup that stores single-photon-level light for tens of seconds, and the retrieved signal after 42 seconds. Figure: Sa et al., Quantum Frontiers 5, 14 (2026), CC BY 4.0

The review worth bookmarking (Journal)

Sa, Liu, Zhou and Li (USTC) published "Long-lived quantum memories with rare-earth ions doped in crystals" in Quantum Frontiers 5, 14 (2026), open access (journal, arXiv:2611.04526).

The most useful thing in it is a strict taxonomy of three lifetimes, ordered so each bounds the next:

  1. Spin coherence time - how long the spin ensemble stays phase-coherent.
  2. Light storage time - how long an optical signal can be stored in the spins and retrieved coherently.
  3. Quantum storage time - how long a single-photon-level signal survives with SNR high enough to beat the best classical strategy.

Confusing these is the root cause of most inflated quantum-memory headlines.

The consolidated records:

Metric Value System Year
Spin coherence ~6 h (370 +/- 60 min) Eu:Y2SiO5 2015
Light storage 1/e = 52.9 +/- 1.2 min, coherent at 1 h Eu:Y2SiO5 2021
Quantum storage (single-photon level) 1/e = 27.6 s; 88.0% qubit fidelity at 5.6 s (classical limit 82.1%) Eu:Y2SiO5 Lv et al.

Eu:Y2SiO5 is still the only material with hour-scale coherent light storage. Requirements for context: repeaters need ms-to-minutes depending on architecture; transportable memories need hours.

The review is blunt about blockers: the lifetime/efficiency doping trade-off, DD pulse-error accumulation and heating at the thousands-of-pulses regime, and integration engineering.

Hype check

A popular outlet ran "hour-long quantum memory breakthrough" this week. It is a writeup of the review above: the hour-scale results are from 2015/2021, not new experiments. Real records, no new result.

Er:YSO ZEFOZ operating points (Preprint)

arXiv:2611.08396 (Nov 5, single author, modeling only):

  • Full-Hessian spin-Hamiltonian treatment of site-2 167Er:Y2SiO5 ZEFOZ points, with anisotropic 89Y-bath field covariance derived from the crystal lattice.
  • Pareto-optimal frontier of four transition pairs balancing modeled pure-dephasing time vs microwave control authority; best modeled Y-bath T_phi > 2.2e5 s; best addressability 1.248e5 Hz/mT Rabi coefficient.
  • ZEFOZ points remain recoverable by numerical recalibration under Hamiltonian perturbations (84.1st-percentile relocation 10.95-27.97 mT).
  • Caveats stated by the author: timescales are modeled and conditional on the temporal-correlation model; assumes ~10 mK operation.

Groundwork, not a measurement.

Also circulating

Last week's ICFO Pr heralded-memory preprint (arXiv:2609.02579) got popular-press pickup; no new data.

Full article with per-paragraph sources: https://alfiomus.blogspot.com/2026/11/this-week-in-rare-earth-quantum.html

Top comments (0)