How Moore’s Law Died or the Evolution of Transistors: From 2D to 3D

Moore’s Law is a controversial statement that emerged in the middle of the last century. It marks the emergence of a trend towards the enormous reduction of electronic components in computer design. Although it has been many years in the making, this trend is still the subject of heated debate.

The origin of the law

The law takes its name from its originator, Gordon Moore. He was the director of the research department at Fairchild Semiconductor. The April 19, 1965, issue of Electronics magazine contained an article by Moore which recounted the results of his research as well as his own observations.

Gordon speculated about the number of electronic components present in integrated circuits. In his opinion, their number increases twofold every year (12 months). To confirm his hypothesis, Moore provided a graph. It shows a forecast of the further development of microelectronics. He took the next decade as an example.

Looking at the graph, it is difficult to identify any valuable information, as the amount of information is minimal. The public’s interest was piqued by the caption that accompanied the image. It read: “Extrapolation of the optimum number of components on an integrated circuit at the lowest possible cost”. It was this phrase that was subsequently corrected several times and became known as Moore’s Law.

Variations in the interpretation of the law

The original “signature” of the graph has undergone changes many times. There are 2 known versions. The first one reports a more reduced component cost factor. The second edition is more common, but omits such a point altogether.

Moore made his suggestion to the public a little over 5 years before the first integrated circuit, known as the Intel 4004, was created. Its chip was equipped with 2,300 transistors. As it turned out, Gordon’s prediction didn’t come true. Therefore, in 1975, he decided to adjust it. Now the 2-fold increase was not for every 12 months, but for 24 months.

The relevance of the law

In order to understand whether the law worked or died out, it is necessary to analyze the hypothesis of Moore with what we have now. To make it clearer, we should take the figure he predicted – 64 components in 1965, as well as the number of components present on Intel 4004.

By refining the data for each year, we can conclude that the starting forecast was significantly underestimated. Interestingly, in the case of the Intel 4004, Moore’s Law worked almost perfectly.

At the time of Gordon’s article the concept sounded convincing enough. It suggested a decent level of efficiency, and that in turn promised good prospects. The concept was accepted by all microelectronics designers, and is also often used as a basic strategy for achieving competitive success in the industry.