Moore’s Law is not just for semiconductors,




Moore’s Law is not just for semiconductors, March 12, 2013

Mathematical calculations of the cost of technology development have shown that the growth of industrial production and performance in general can be predicted in many areas, including aviation. Previously responsible for it except that Moore’s Law, or rather not answer, and predicted the various restrictions, but it turns out, the development of semiconductors – is not the only patrimony of Moore’s Law. Moreover, no one ever bothered to check the operation of Moore’s Law. Up to the present moment.

Prediction of future technologies like playing the fool. In 1946, for example, Thomas J. Watson, the founder of International Business Machines, now known as IBM, argued that the world will only need five computers. We have already written to you about how sometimes diverge predict even the next five years with reality later these same five years, but now let’s talk about something else. U.S. researchers report that today the technical progress is very predictable – and provide evidence against 62 different technologies.

Like we have already heard. But a group of researchers at the Santa Fe Institute in New Mexico, in conjunction with the Massachusetts Institute of Technology in Cambridge, decided to verify the foregoing to the full. In a study published in PLoS ONE, compared to a set of mathematical laws, which apply for a description of how the costs of developing technology. Scientists have found that the most accurate of these laws was launched in 1936.

This law brought aeronautical engineer Theodore Wright (Theodore Wright), who noted that the cost of the plane falls to the extent that it increases the number of aircraft produced. In particular, the cost is inversely proportional to the number of planes that are getting better. This theory has been something of a universal law, to regulate the cost of technological products, and is often interpreted in a simple way: the more we do, the better and more effective is our manufacturing process.

However, much more is known law, derived in 1965 by Gordon Moore, co-founder of the production of microelectronics Intel. He noticed that the power of the computer for the dollar is growing exponentially for quite some time. In essence, this means that the cost of one semiconductor decreases exponentially.

Offered to other relations between the scale and the cost of production: for example, that the costs are falling only because of the scale of production. All these laws predict that costs will fall over time – but differ in the details.

“Until now, these theories no one checked” – govort Trensik Jessica (Jessica Trancik) from MIT. She and her colleagues collected data on 62 technologies from chemicals to electronics (for example, solar cells) and Information Technology, created over a period of 10 to 39 years. – “Collection is big enough, and score data was very difficult.”

The researchers evaluated the effectiveness of six such laws estimation – took the initial data, and tried to predict the subsequent costs – and then evaluated as they relate to the real data.

It turned out that the laws are not particularly different. The most accurate was the law of Wright, but Moore’s Law behind quite a bit, at least, within a few decades. Predictions so clearly combined with these two laws, the researchers determined that the laws, in turn, may be closely related.

Connection is very clear. In 1979, political scientist Devendra Sahai (Devendra Sahal) noted that if the production of hardware units is growing exponentially, the laws of Wright and Moore equivalent. The data confirm that the production is really growing exponentially, if sufficiently broad range of products.

Suddenly it became clear that Moore’s law is valid in a large range of industries that computer technology is only considered as a special case.
Market forces

Economist William Nordhaus of Yale University in New Haven, Connecticut, warns that almost always these laws work in the technologies to survive, so we can not predict the future growth is very young technology.

“History is written by the winners,” – he said. – “Those technologies that do not go off the market, do not fall into the list. This is one of the main reasons why we can not predict for sure which of the emerging technologies in the energy sector will survive. ”

The future of some of the technologies depends on the overall market, and not only on its fluctuations. For example, the evolution of technology, the changing climate, which specializes Nordhaus, will depend on future pricing policies on carbon emissions. As the economist, “some technologies, such as the capture and storage of gas uglekislovogo not shoot if the market price of carbon will be zero.”

Assessment of the potential costs in technology, climate change, is the cornerstone on which the researchers are based in their design. Speaking of climate change: the latest data from Antarctica confirm that the ice is melting twice as fast than before.
Higher intelligence recommends:

5 myths about the technology of the future
Samsung took away the title of Apple’s largest consumer of semiconductors
Car DVR HD Defender Car Vision 2020HD

Tags: Moore’s Law , Semiconductors, Economy



Tags: , , ,

In: Technology & Gadgets Asked By: [ Grey Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »