On Mon, 25 Mar 2002, Richard A Steenbergen wrote:
> On Mon, Mar 25, 2002 at 11:05:21PM -0500, Deepak Jain wrote: > > > > That is a falicy. Moore's law is most certainly not accelerating -- in > > fact: > > > > 1965-1990 Moore's law stated that the number of transistors per square > > inch on integrated circuits (and therefore, the speed) doubles every 2 > > years. The pace has since slowed down a bit, but appears to be holding > > steady at doubling every 18 months (1995-present). > > Not to be too picky, but how is going from "doubling every 2 years" to > "doubling every 18 months" slowing down? :) Erm, yeah. Thanks for calling me on that -- I horribly condensed what I was trying to say. By the original definition (number of transistors per square inch doubles every year), it has slowed to every 2.5 years. See the graph I linked to. Data density is currently doubling every 18 months, and holding steady at that rate. (But this *is* off topic for NANOG...)