January 29, 2007 | posted by Michael Baxter in IABN |

The main difference between the underlying principle of Moore’s law, and Vicky Pollard, that character from Little Britain, is “but.”

Ms Pollard would say, “yes, but no, but”. Moore’s Law is just about yes and no, or on and off. Each transistor sitting on a chip is effectively saying yes or no, and the particular selection made determines whether you get a zero or a one. Such are the building blocks of digital technology. With digital, there are not buts.

The more transistors you fit onto a chip, the faster it can process information. Back in the 1960s, Gordon Moore predicted a doubling in the number of transistors on each chip every 18 months to two years. But, of late, many have feared we are approaching the end of this period of ever faster chips, that they were getting too small to work with.

The smaller the chip, the higher the level of energy loss, which translates into heat. Scientists have feared that if chips get much smaller they will simply be too hot for their technology too handle.

Now, both IBM and Intel have separately announced new technology for, partially at least, solving the problem. Some experts are calling it the most significant breakthrough in forty years.

The technology, which was developed by both companies independently of each other, uses something called high-k material. The result, according to Intel, is a transistor that will require 30 percent less power for it it to be switched on and off. Apparently, the new Intel chips - to be launched in 2008 - using the technology will be just 45 nanometers. To put that in context, a nanometer is just one billionth of a metre thick.

As scientists pour their energy into reducing the heat created as chips gets smaller, the battle between AMD and Intel also heats up. And despite the huge research bill Intel will have incurred in developing this latest breakthrough, the advantage has not necessarily been ceded to the larger chip company. Apparently, AMD has been working with IBM on its RD, so will also have access to similar technology.

Last summer, we revealed news from Intel on another technological breakthrough for helping maintain adherence to Moore’s Law.

The company announced a breakthrough with a substance called Indium phosphide. The theory works like this: Silicon chips need to get smaller so that the information between each transistor can be passed more quickly. But if a way could be found for sending this information at the speed of light, the chips would not need to be so small.

Silicon is fine when it comes to detecting, modulating and even amplifying light, but it can’t create it. Indium phosphide, however, can. And the Intel scientists reckon they have found a way of getting the two materials to work together.

So, it seems Scientists are having their cake and eating it; having the cake of being able to build smaller chips, while eating the cake of not needing to make the chips smaller anyway.

It’s a case of little and large: either way does it.