Gordon Moore’s 1965 observation of increasing integrated circuit power paralleling shrinkage in size was originally tentatively phrased: “The complexity for minimum component costs has increased at a rate of roughly a factor of two per year … Certainly over the short term this rate can be expected to continue.” And yet it’s held more or less true for forty years, and new research suggests that with continued innovation there’s plenty of life in the Law yet.
Moore’s law in “strict” interpretations argues that transistor density on ICs doubles every two years, and thus far this has been achieved with continuous innovation in chip fabrication technology, adjustments to the semiconductor materials used, and improvements in chip architecture. But as transistor sizes have shrunk down towards 20nm–Intel today announcing its 32nm process–hard physics comes into play. Both quantum mechanical effects and heating difficulties mean that practical transistors based on existing semiconducting materials much smaller than 20nm will simply be too unreliable to work in chips.
But research published by the National Physical Laboratory in the UK has pointed out a new direction for ICs: magnetic semiconductors. Specifically the team’s been working on germanium nanowires doped with manganese, and shown that they have excellent potential to act as nano-scale transistor components. This innovation of magnetic semiconductors–which don’t occur in nature–and the fact that they’re potentially highly compatible with existing silicon-based circuitry means that if the technology is commercialized into chip fabrication then Moore’s law will likely hold for a few decades more, and future chips will be yet more powerful and less power-hungry. And more processing punch with less battery-drainage makes for better consumer products.