Chapter 2: Trends Towards Mobility 31 ization. The famous “Moore’s law” by Gordon E. Moore (1965) estimates that computing capacity (CPU speeds, memory sizes) will double every 1.5 years or so. This astonishing prediction has held true for a long time, largely thanks to the increasing miniaturization of inte- grated circuits. The diminishing size of transistors allows for larger numbers of elements onto chips and decreased power consumption. As chip makers cram ever more elements on chips, new functions become practical, and old functions become faster – or use less power. A similar development, thanks to advances in micromechanics and quantum physics, has allowed hard disk manufacturers to build ever denser storage devices. At the turn of the century, hard disk capacity was increasing faster than the Moore’s law predicted, due to a leap in magnetic reading head technologies. At the time of writing, the capacity race has slowed down, but mass storage prices keep falling, so the amount of storage per unit of money still follows the Moore’s law. There have been, over time, predictions of Moore’s law failing due to various physical barriers. For instance, as microcircuit fea- tures sizes decreased below 100 nanometres, the inherent uncertainty of manipulating electrons in near-atomic size circuits was expected to render further improvements in digital electronics impractical. However, advances in solid state physics, circuit design, and IC manu- facturing have made the current nanoscale processes an everyday practice. As another example, the ever increasing number of input/output connections of the chips demand more pins on their packages, which makes device design more diffi cult. To counter this diffi culty and gain yet more speed, major processor makers are increasing the number of CPUs on a single chip to gain performance by multiprocessing, and integrating peripheral devices (such as cache memories) on the same chips. It appears that the various physical barriers will eventually be cir- cumvented through advances in physics, clever circuit design, and better manufacturing. However, we are not expecting fundamental slowdowns in the advance of electronics in the next decade or two. Technology will march on – because it can. 2.5 Convergence So the trend of better, smaller, cheaper, faster will continue. What will it give us?