Flattening the Curve
Computerworld ran an interesting piece last week about the slowdown in the phenomenon known as Moore’s Law.
Moore’s Law was never a law in the scientific sense. Gravity is a law. What Gordon Moore postulated in 1965 was an exponential growth curve that said the number of transistors that engineers could put on a chip would double and the cost of computing would halve about every 18 months for the foreseeable future. It was a no brainer almost a tautology, really. It lasted only 50 years and just when things started to get interesting, bang!
The trouble with growth is that some factor always intrudes to stop it. Food limits population size unless you can keep expanding the supply for example. In any system with a finite resource base infinite expansion is impossible to contemplate. So it is with computer chips. There is only so much real estate on a chip.
According to the article chip designers are working at the nanometer level with only a few atoms between transistors now so it is hard to see how the next doubling is going to happen. More expensive processes and materials can put off the inevitable but by then you’ve been painted into a corner.
This does not mean that computers can’t get any smarter or that we are at some limit to our software creativity. Bigger chips, better strategies for chopping up hard compute intensive jobs can still make a difference. But what will this mean for the size of, say, your handheld device? Think of it as a carton full of eggs, where are you going to put the next one? You can make a bigger box but how big can you make a chip?
You might call this a limit on growth and this has been much on my mind since I read the 30-year update on the original publication — “Limits To Growth” — from The Club of Rome recently. That cheerful ditty applied computer modeling to a host of data and concluded that at 8 or 9 billion people on the planet, with increasing pollution, decreasing energy resources and raw materials, and decreasing food production, things get a little dicey.
Don’t look now but with increasing population, global warming, droughts or heavy rains affecting farmlands, depleted fishing grounds, expensive energy, and declining raw material bases, we seem to be on schedule. You can’t expect infinite growth with finite resources. First Moore’s Law, then what?