Is there a Moore's Law of the Blogosphere?
The reason for asking that question is the announcement this week by blog tracker Technorati (a great site, by the way, for continuously following the state of the Zeitgeist), in its annual State of the Blogosphere report that the number of blogs in the world has jumped from 7.5 million in March to 14.2 million today.
In other words, in appears the blogosphere is doubling in size every five months. Or even more staggering -- a new blog is being created out there somewhere every second.
Whenever you hear the word "doubling" related to anything high tech, the first thing that comes to mind is the Law of Laws in the digital world: Moore's Law of Semiconductors. I probably don't have to remind you of what it says: the performance of semiconductor devices doubles every two years.
Gordon Moore, one of the pioneers of Silicon Valley and a co-founder of both Fairchild and Intel, came up with this "law" in the mid-1960s while preparing for an industry speech. Basically, he tracked the capacity of all of his company's memory chips up until that date … and discovered, to his astonishment, that they made a straight line on a sheet of log paper. He then took a chance and predicted that this trajectory might be maintained for a few more years.
I saw Gordon make a follow-up presentation in the late 1970s. He was still amazed: the chart now included logic chips, memory chips and even microprocessors -- and yet the semiconductor world was still clicking along at this mind-boggling pace, doubling performance every 18-24 months. And it worked along multiple axes: hold performance constant, and chip prices halved every couple years. Hold those two variables steady and chips got smaller at the same rate.
By this point Moore would only say that he could no longer predict how long his "law" might last, but surely it would have to eventually crash into some unbreachable wall: fabrication cost (price) or quantum physics (miniaturization) or mathematics (perpetual doubling, like the proverbial grains of rice on a chessboard, would eventually approach infinity).
Yet, when I last talked with Gordon a couple years ago about the law, which has now made his name immortal, he could only chuckle in disbelief that not only was it still at work, but showed no sign of slowing down. The Semiconductor Industry Association, which he advises and which prepares the biannual Industry Roadmap, now sees no serious obstacles to Moore's Law until at least 2015, and perhaps well into the middle of the century. By then, $50 microprocessor chips will have greater processing power than the human brain, and home PCs will be as powerful as today's biggest supercomputers.
As has been rightly noted, Moore's Law isn't really a law, but a contract between the semiconductor industry and mankind to continuously drive chip technology forward at the greatest possible pace. In that respect, it has become largely self-fulfilling: the world now expects a certain rate of change, and the semiconductor industry struggles mightily, generation after generation to reach that goal.