Group items matching

in title, tags, annotations or url

Between 2006 and 2010, demand for processing cycles (MIPS, servers and the like) has slowly approached an 18 erpcent annual growth rate in the big banks. Storage, by the way, has hit 45 percent per year -- the advent of Big Data is here -- and although the unit cost of storage is still dropping, storage cost pools around the financial industry are expanding out of control. The growth phenomenon is now exacerbated by market conditions, and Moore's Law just isn't enough.

Taking a step back, you will likely ask, "How can this be true?" The answer involves yet another "law" -- actually, a paradox observed in the late 1800s -- "Jevons paradox," which states:Technological progress that increases the efficiency with which a resource is used tends to increase (rather than decrease) the rate of consumption of that resource. William Stanley Jevons developed this hypothesis in 1865, based on his observations of coal consumption vis-à-vis the technology advances designed to improve the efficiency of coal usage. It was his argument that these improvements alone could not be relied on to reduce consumption; rather, they would lead to increased consumption -- and he was right. Today we talk about elastic computing; in 1865 Jevons focused on "elastic coal" – well, at least the demand was elastic.

So the aforementioned growth in demand (passing the 20 percent mark per year) is actually fueled in part by the inherent efficiencies created by Moore's Law. Through 2010 we were in the Moore's Law zone of managing IT costs downward. Now we are a new world governed by the effects noted by Jevons.