http://phys.org/news/2014-09-non-volatile-memory-energy-efficiency-magnitude.html
what surprises me about this is that this huge improvement in computer memory energy efficiency doesn't involve spintronics although it should be possible to combine this with spintronics on the same circuit. I would have thought that, for now on, any such big orders-of-magnitude energy efficiency improvement in microcircuit electronics would almost bound to involve spintronics because that is surely the future for ultra energy efficient microcircuit electronics.