Since we seem to have no other choice now except be slaves to Moore's law (or accept that computers and Iphones aren't going to get faster and smaller at quite the same rate as before), researchers have to find ways of codifying information into smaller and smaller materials. Current microchip technology uses arrays of semiconductor transistors whose current can be turned "on" and "off" using an externally applied electric field. This on/off switch is what makes the ones and zeros that are the building blocks of computer programming languages. However, companies like Intel are starting to reach the limits of how small they can pattern these semiconductors, which restricts how many transistors they can build into a microchip of a given size. What's crazy is that these transistors are already just tens of nanometers wide, but the next leap forward in computational density requires researchers to develop angstrom-scale materials which display similar switch-like properties,
Enter the atom.
Atoms certainly put nanotechnology to shame in the size department and they may even have the ability to act like a switch for our greedy computing demands. Anyone who has muddled through quantum mechanics can at the very least tell you that an atom's spin produces a magnetic moment, and like all vectors, that moment has a direction. The direction of the magnetic moment can be flipped to align with an applied magnetic field, but it will relax back after that field has been removed.
What the researchers in Donati et al. showed was that under certain conditions (cryogenic temperatures for one), they can isolate a Holonium atom on a metal oxide surface, switch its magnetic direction, and maintain it for as long as 1500 seconds after the field has been removed. That's a huge step forward in magnetic stability, or remanence, which could be used to build a magnetically driven switch; perhaps one day transforming a humble Holonium atom into a very attractive one or zero.
As a former carbon nanotube researcher, I'm personally interested to see whether the semiconductor manufacturing industry will even mess with graphitic materials in future iterations of the microchip, or whether they're going to skip that idea altogether and move straight into paradigm shifting technology like single-atom computing. What do you think they'd call themselves then? The atomic manufacturing industry? Seems inevitable.
You can find the original Donati et al. article here, in Science.