NIST's new, even more precise atomic clock

One wouldn't think that being a second off every, oh, 70 million years or so, would be such a huge deal, right? Apparently that benchmark just isn't timely enough for the National Institute of Standards and Technology, whose Time and Frequency Division has fabricated an experimental atomic clock based on a single mercury atom in favor of the fountain of cesium atoms used now. They've discovered the prototype is even more accurate than the current standard, and would only lose one second every 400 million years. Obviously nobody reading this will even be around in 400 million years (um, right?), but there are reasons to improve aside from holding the time steady: precise time-keeping aids in accurate syncing of GPS and navigation systems, telecommunications, and deep-space networks. We admit, this whole thing leaves us a bit flabbergasted, but the sense of absurdly painstaking scientific security we'll get from knowing that while civilizations rise and empires fall, no one will live to see this atomic clock miss a beat -- well, that couldn't have come a moment too soon.

[Product Reviews @ Jul 17th 2006 7:03AM]

I love the way tech things keep getting better, but is this really needed as I am not sure how many of us will be around in 70 million years or so.

[Dave @ Jul 17th 2006 7:17AM]

Loosing a second every 70 millions years may not seem like a big deal to a lot of people, but really it is.

Where I work we have several cesium clocks to measure time dialation and distortion (Mainly. They are used for loads of other things as well). They are used to measure incredibly small discrepancies between two synced clocks.

These clocks have to accurate. Infact, nearly damned perfect or you'll get an incorrect, or worse yet, wrong result.

This is great news as our most accurate (and expensive!) cesium clocks will no longer be good enough for the job in about 10 years.

[RacetrackOwner @ Jul 17th 2006 7:28AM]

How far off the clock is in 400 million years is just a side-effect of the reason a more accurate clock is required. It's kind of sad that this point seems to be lost on so many people who are supposedly interested in high tech news.

[Peter McCaffrey @ Jul 17th 2006 7:43AM]

mmm, the point is that it makes measuring really SHORT times more accurate.

[Bushy @ Jul 17th 2006 8:10AM]

What's the difference between an "incorrect" result and a "wrong" result...?

[Dave @ Jul 17th 2006 8:24AM]

"What's the difference between an "incorrect" result and a "wrong" result...?"

Sorry, incorrect should of course be inaccurate. But I didn't need to tell you that though, did I?

[Stephen @ Jul 17th 2006 8:29AM]

An "incorrect" answer will get you fired; a "wrong" answer will get you (and probably other people) killed.

[tristanfey @ Jul 17th 2006 1:03PM]

As has been stated above, the importance of having a clock this accurate is its usefulness in measuring very small amounts of time. For example, access time for RAM is measured in nanoseconds (ns). A ns is only 1 billionth of a second or 10 to the negative 9th power.

Measurements in laser technology require the use of femtoseconds (10 to the negative 15th power), which are a millionth of a ns.

And if you are a photon researcher you need to make measure in attoseconds (10 to the negative 18th power), which are a quintillionth of a second.

[blackfeather @ Jul 17th 2006 1:10PM]

With accuracy like that, you've probably got to be careful how you transport it. Move one of them too fast, and relativity will pull them out of sync. They've probably got to compensate for that if they're used in GPS sattelites.

[Jeff @ Jul 17th 2006 1:21PM]

Actually the GPS satellites already compensate for relitivety. Relitivity accounts for about a 38 microseconds per day error, so to compensate they run the frequency off the satillite a little slower than it would be on earth, 10.22999999543 MHz instead of 10.23 MHz, which through what we in the aviation field call PFM (Pure F---ing Magic) corrects the error.

[Rick @ Jul 17th 2006 2:15PM]

Better atomic clocks, better accuracy. Build me a GPS with 1" accuracy please!

[Windell Oskay @ Jul 17th 2006 9:09PM]

The picture shown is not actually of the mercury clock; it's of a different clock, built by a different team. (Blame Roland Piquepaille, he mistakenly included that picture in his blog entry here: http://blogs.zdnet.com/emergingtech/?p=293 )

[Jeff @ Jul 18th 2006 1:07PM]

Rick,
Actually they have GPS technology that is accurate to within 1 cm. They use a technique called RKP, Relative Kinematic Positioning, to correct GPS error to within 1-10 cm. I don't know of, and would highly doubt it's availiblity to the public....yet.

[Andy @ Jul 20th 2006 8:48AM]

Actually, it's not only for measuring short times, but also for measuring small differences in time lapses, which is not the same. This is how the theory of Relativity was agreed by experiment for the n-th time.

Also, take into account that any experimental evidence of physical theories, be it quantum mechanics, relativity, or M-theory (OK, potential evidence too, especially at the quantum level, is only proven by certain margin of error, which happens to be the most accurate margin of error in any scientific theory, for quantum mechanics, due to quantum uncertainty. This minute, almost non-existent margin of error could get even smaller with these advances.