- A rule that predicts the number of transistors on a circuit board has helped shape the entire technology industry for 50 years.
- Using Moore’s law, smartphones and self-driving cars were predicted decades in advance.
- Physical limits mean Moore’s law may now be losing relevance.
We are all familiar with Murphy’s law, the popular proverb that claims the worst outcome is the one most likely to occur. A similar-sounding (but arguably much more important) adage is Moore’s law, which for decades has steered the rapid development of the world’s computers and electronics. And that law is about to be broken.
For more than 50 years, Moore’s law has been used to predict the technological advancement of the world’s computers. Without it, there would be no mass consumer electronics like laptops, smartphones or wearable technology, no robotics or smart machinery, nor the data centres and servers that power Google, Facebook, or even the world wide web as a whole.
Nonetheless, as the world’s mainframes and microchips continue to advance, the law is beginning to reach the upper limits of its predictive power. As the New York Times noted earlier this month, it has gotten to the point where one of the computer industry’s most important governing bodies is proposing a new framework that will eventually replace Moore’s law altogether.
law and order
Essentially, Moore’s law says that the number of transistors on a circuit board doubles every two years. It is named after Gordon Moore, an American electrical engineer and businessman who co-founded Intel Corporation in 1968. Three years earlier in 1965, Moore first made these breakthrough observations in the wonderfully-titled paper Cramming more components onto integrated circuits.
Using this roughly exponential formula, consumer electronics designers and manufacturers could predict, with a fair degree of accuracy, what technological capabilities would be unlocked or made more affordable in the near future. This allowed for upcoming products to be conceived and designed well ahead of time, despite their most crucial internal components not yet existing.
A good example of the capabilities of Moore’s law comes from Moore himself. In the opening passage of Cramming more components, Moore predicted that the acceleration (and associated miniaturisation and falling costs) of computing power would lead to a wide range of new gadgets. Gadgets that, until then, existed only in science fiction.
‘Integrated circuits will lead to such wonders as home computers… automatic controls for automobiles, and personal portable communications equipment,’ he wrote, uncannily foreseeing the rise of cell phones and self-driving cars decades before their respective appearances.
It was not just microprocessor chips that were subject to the law. Hard drives, memory sticks, and other crucial components also featured a similar growth curve. You could even argue that a company such as Instagram is a product of Moore’s law, as its business model depended on the monumental improvement of smartphone camera technology when it launched in 2010.
While Moore’s law held mostly firm for the rest of the 20th century, it began to waver in the new millennium. In a 2005 interview, Moore himself said the law ‘can’t continue forever’, due to a tendency for exponential models to plateau.
In the case of computers, the challenge is literally physical. As more and more circuitry is packed onto smaller and smaller matter, computing must move from the microscopic level to the atomic level or beyond if the same rate of growth is to be maintained for much longer.
Recognising this fundamental problem, the Institute of Electrical and Electronics Engineers (IEEE, the world’s largest association of technical professionals) announced in early May that it is launching the International Roadmap for Devices and Systems (IRDS), a new method for charting the growth of the computer industry.
By moving beyond integrated circuits and incorporating other elements such as software and systems architecture, the IEEE hopes the IRDS will pick up where Moore’s law left off.
they stay the same
While things look mighty grim for Moore’s law, its ability to stick around means it shouldn’t be completely discounted. In an April announcement, Intel CEO Brian Krzanich staked his company’s future plans on its continued efficacy, noting that he had “witnessed the advertised death of Moore’s law no less than four times,” throughout his career.
Meanwhile, other cutting-edge concepts such as quantum computing, which does away with transistors and instead seeks to harness the changing states of matter itself, indicates there is still much room for innovation in the computer technology industry.
Whatever happens, it is a fairly safe bet to back the continued growth and advancement of computing power, whether these evolutions are subject to Moore’s law or not. And if the model is eventually done away with altogether, hopefully this will not result in a monumental case of Murphy’s law.