- NIST confirmed that several public time servers lost their atomic reference signal
- A generator failure interrupted the distribution of America’s primary atomic time scale
- Some NIST servers responded normally while quietly serving inaccurate timestamps
The National Institute of Standards and Technology (NIST) has issued a warning that some of its public time servers may be unreliable.
The advisory focuses on a defined set of hosts, including multiple time-xb.nist.gov addresses and the authorized ntp-b.nist.gov service.
According to NIST, these systems can still respond to network requests while no longer referring to a valid atomic time source.
What went wrong at the Boulder facility
To avoid spreading incorrect data, the agency said it may temporarily take some of the affected hosts offline.
NIST traced the problem to its Boulder, Colorado campus, where an extended power outage interrupted operations.
The outage occurred during high winds that damaged power lines and triggered safety-related shutdowns.
Although backup power systems were in place, a downstream generator failure interrupted the atomic time scale distribution that powers the Internet Time Service.
NIST stated that the UTC(NIST) signal drifted by about four microseconds during the event, a deviation that is small but measurable.
The outage does not affect all NIST endpoints. Widespread addresses such as time.nist.gov rely on round robin DNS and geographically distributed infrastructure.
This design allows clients to automatically fall back to unaffected sites when a site encounters problems.
Users who hardcode individual hostnames face greater exposure to localized errors like this.
Systems running on cloud hosting platforms often rely on pooled or upstream time sources, which can mask short-term problems at any single facility.
The Boulder site hosts the NIST-F4 atomic clock, which uses cesium atoms to define the length of a second with extreme precision, supporting services used by telecommunications networks, power grids, financial platforms and scientific research.
Accurate timing is also fundamental to data center hosting environments, where synchronization affects logging, security protocols, and transaction ordering across distributed systems.
Many enterprise servers trust external authoritative sources, making upstream accuracy a common dependency.
This incident follows another timing service outage earlier this month at NIST’s Gaithersburg, Maryland, facility, which caused a major time step error measured in milliseconds, not microseconds.
NIST has not given a firm timetable for full restoration of Boulder, and said engineers are continuing restoration work.
While most consumer systems will hardly notice the problem, high-precision users are expected to monitor multiple independent references.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds. Be sure to click the Follow button!
And of course you can too follow TechRadar on TikTok for news, reviews, video unboxings, and get regular updates from us on WhatsApp also.
