It may not be as accurate as today's atomic clocks, but it's an awful lot simpler in concept: a new way of measuring time based on the mass of an atom.
Taking advantage of the fact that matter can be both a particle and a wave, Holger Müller, an assistant professor of physics at the University of California, Berkeley, has found it's possible to tell time by counting the oscillations of a matter wave. A matter wave's frequency is 10 billion times higher than that of visible light.
Müller refers to his method as a Compton clock because it is based on the so-called Compton frequency of a matter wave.
"When you make a grandfather clock, there is a pendulum and a clockwork that counts the pendulum oscillations. So you need something that swings and a clockwork to make a clock," says Müller.
"There was no way to make a clockwork for matter waves, because their oscillation frequency is 10 billion times higher than even the oscillations of visible light."
The solution came through combining two well-known techniques. In relativity, time slows down for moving objects, so that, for example, a cesium atom that moves away and then returns is younger than one that stands still.
As a result, the moving cesium matter wave will have oscillated fewer times: and the difference frequency, which would be around 100,000 fewer oscillations per second out of 10 million billion billion oscillations, turns out to be measurable.
The Compton clock is still 100 million times less precise than today's best atomic clocks, but could perhaps be improved to the same level.
The technique can also be turned around to use time to measure mass - possibly one day replacing today's one-kilogram reference mass, a platinum-iridium cylinder. Just this week, researchers warned that this had become a little overweight, thanks to contamination.
"This is a beautiful experiment and cleverly designed, but it is going to be controversial and hotly debated," said John Close, a quantum physicist at the Australian National University in Canberra. "The question is, 'Is the Compton frequency of atoms a clock or not a clock?' Holger's point is now made. It is a clock. I've made one, it works."