Science history: Computer scientist lays out ‘Moore’s law,’ guiding chip design for a half century — Dec. 2, 1964
Milestone: Introduction to Moore’s Law
Date: December 2, 1964
Or: San Francisco Bay Area
WHO: Gordon Moore
In a low-key lecture to a local professional society in 1964, computer scientist and chemist Gordon Moore outlined a prediction that would define the world of technology for more than 50 years.
The final version of this prediction would be known as “Moore’s law“, and it would drive progress in the semiconductor industry for decades.
Even though it’s called a law, it was a prediction based more on economic dictates and industrial trends than the physical laws of nature.
Moore was director of research and development at Fairchild Semiconductors when he gave his speech, and his goal was to ultimately sell more chips. At the time, computers were gigantic machines that took up an entire room, and integrated circuits, called microchips, had somewhat limited practical applications.
The silicon transistor, the workhorse that performs the calculations in a computer, had been invented barely a decade earlier, and the integrated circuit, which made it possible to miniaturize computers, had been patented just five years earlier. In 1961, electronics company RCA had built a 16-transistor chip, and in 1964, General Microelectronics had built a chip of 120 transistors.
Moore witnessed this dramatic progress and noticed that a mathematical rule seemed to govern this progress. This mathematical correlation was later called “Moore’s Law” by other people.
Although Moore outlined the principle to the Electrochemical Society in 1964, he gained traction in April of the following year, when he was asked to write an editorial in Electronics magazine. In it, he boldly predicted that until 65,000 components could be grouped on a single chip — an unheard of figure at the time. That’s a lovely, small number now, considering that in 2024, a company revealed a 4 trillion transistor chip.
In 1968, Moore would co-found chipmaker Intel, where his law of twos would evolve from a simple observation to a motivation for innovation.
Despite its name, Moore’s Law has never been a foolproof rule. In 1975, Moore reduced the rate of progress until doubling the transistors every two yearsrather than every year. This more modest doubling rate would become official Moore’s Law, which would remain valid for years. This relentless trend toward more computing power and miniaturization is behind virtually all modern electronic devices, from the personal computer to the smartphone.
For years it was predicted that the law would become obsolete, but it proved remarkably resilient for a time.
“The fact that we were able to continue [Moore’s law] so long surprised me more than anything,” Moore said in a interview with The Electrochemical Society in 2016. “There always seems to be an impassable barrier in the road, but as we get closer to it, people are finding solutions.”
But ultimately, the principle would no longer hold. It is not clear when Moore’s Law disappeared. In its canonical form, the norm probably died in 2016because it took Intel five years to move from 14 nanometer technology to 10 nanometer. Moore saw this happening, as it was years before his death at the age of 94 in 2023.
Eventually, Moore’s “law” must have run out of steam because it clashed with the real laws of physics. As transistors got smaller and smaller, quantum mechanicsthe physics which governs the very small, began to play a disproportionate role. The world’s smallest transistors can facing problems with the “quantum tunnel”” in which electrons from one tiny transistor can flow into another, allowing current to flow through transistors that should be in the “off” position.
As a result, chipmakers are considering designing chips with new materials and architecture. Moore’s next law could apply to quantum computerswhich exploit quantum mechanics as a feature and not as a calculation bug.
:max_bytes(150000):strip_icc()/Health-GettyImages-1390051226-f85d922918fb4c77a5db2c64b2386259.jpg?w=390&resize=390,220&ssl=1 "10 Reasons You Aren’t Feeling Hungry")
