Moore's Law: Did He Predict the Future?

Moore's Laws are empirical laws. They relate to the evolution of computing power and complexity of hardware. Strictly speaking, we should not talk about Moore's Laws but as "Moore's guesses" as statements of Moore are in fact only assumptions that, if they could prove true for some time, are doomed to failure in some medium term.

There are actually three "laws" of Moore, two authentic (as they were issued by Gordon E. Moore), and a series of "laws" which have in common is entitled "Moore's Law" but note that they are inaccurate simplifications.

Moore's Law

One microprocessor (Intel 4004) was invented in 1971. It was a calculation unit of 4 bits, clocked at 108 kHz and integrating 2,300 transistors. The integration capacity of transistors and reducing the heat improved performance of processors.

1. Moore's Law has been expressed in 1965 in "Electronics Magazine" by Gordon Moore, an engineer from Fairchild Semiconductor, one of three founders of Intel. Noting that the complexity of semiconductor proposed entry doubled annually in cost constantly since 1959, the date of their invention, he assumed the continuation of this growth (in 1965, the most powerful circuit included 64 transistors).[1] The exponential increase was quickly named Moore's Law, or on account of adjustments later, Moore's first law.

2. In 1975, Moore revaluated himself by asking his prediction that the number of transistors on microprocessors (rather than simply less complex integrated circuits composed of independent components) on a silicon chip doubles every two years. Although not a physical law but just an extrapolation, this prediction proved remarkably accurate. Between 1971 and 2001, the density of transistors doubled every 1.96 years.[2] Accordingly, the electronic machines have become less expensive and more powerful.[1]

3. A common version, variable and unrelated to the actual statements of Moore is "something" doubles every eighteen months, this thing is "power", "capacity", "speed" and many others variants but very rarely the density of transistors on a chip. These pseudo "Moore's law" are those most often running because they flower in publications and on many websites. Their only common point is this number of eighteen months, yet not found in any of the two statements of Moore.

Now you know the truth about Moore, you can see he really only spoke of computer processors. It will be exciting to see if Moore's Laws holds true in the coming years.

References:

[1] Waldner, Jean-Baptiste (2008). Nanocomputers and Swarm Intelligence

[2] http://www.intel.com/technology/timeline.pdf Intel Corp.