Optical-clock ratio study pairs lower uncertainty with a call for repetition

A revised optical-clock comparison study posted to arXiv on July 16, 2026 reports total fractional uncertainties at or below 3.2 × 10⁻¹⁸ for ratios among aluminium-ion, ytterbium and strontium clocks. The preprint also reports differences from earlier measurements, making repetition a central part of the story.

NIST's April 10 publication record describes the same research. The work first appeared on arXiv in December 2025; the NIST page date is not the date of the first experiment or a separate replication.

Comparing clocks is a network task

The study connects instruments at NIST and JILA through a 3.6-kilometre phase-stabilized fibre link and a shared reference. The result therefore concerns a comparison system as well as the clocks themselves.

This matters when interpreting the uncertainty. A frequency ratio is a dimensionless comparison between standards. Its uncertainty is not automatically the timing error a customer would experience from a distributed service, and it is not a specification for a portable instrument.

The authors identify the measurements as meeting an important criterion on the path toward redefining the SI second. That is a research milestone, not a declaration that the definition has changed.

Why discrepancies belong in the headline

Differences from earlier results do not make the new work uninformative. They help identify why repeated measurements and comparisons by different laboratories are necessary.

Our reading is that increasingly precise instruments raise the value of a transparent comparison record. Readers need the measurement version, uncertainty treatment and reference arrangement, not simply the smallest number in a summary.

The NIST record and arXiv version are references to the same work, rather than two independent confirmations. This article treats the work as a preprint and has not independently recalculated the results.

The next useful evidence is further comparison, including independent laboratories and an explanation of remaining discrepancies. Precision is an achievement; reproducibility is the continuing test that makes it broadly useful.