Engineering Explained: Moore’s Law

November 15, 2020

In the 1960s computers were new and they looked nothing like what we know and use today. Computers of this era filled entire buildings at their largest and were the size of an upright refrigerator at their smallest. But, as time marched on they were to decrease in size—a trend that is still prevalent today. With each passing iteration, the goal was to make computers smaller, faster, and more economical. A major contributor to this reduction in size was due to a little invention called the transistor.

A Transistor is a semiconductor device used for amplifying, controlling, and generating electrical signals, according to Encyclopedia Britannica. Multiple transistors are used to form logic gates, which are the fundamental building blocks of integrated circuits (IC)s. The transistor was discovered at Bell Laboratories in the 1940s and quickly became a viable alternative to vacuum tubes, which were used in all electronics at the time. Transistors took up far less space, consumed less electricity, and we’re far more robust than fragile glass tubes. Once we could produce integrated circuits with thousands and millions of transistors, the computer industry began to take the shape of how we consider it today.

But, let’s not get ahead of ourselves here. Let’s look at the transistor in terms of what a certain Gordon Moore had to do with it. Fresh from the U.S. Navy in 1956, Gordon Moore found himself working at Shockley Semiconductor Laboratory in Palo Alto. You may know the place. That lab was researching these new silicon-based transistors. But, things weren’t as great as they seemed. The engineers didn’t like Shockley’s management style and they weren’t getting the results they wanted. Gordon and some of his fellow engineers were not happy there, so they quit. In 1957 the group of defiant engineers started the Fairchild Semiconductor Corporation in Santa Clara, California. Within two years they were producing their own transistors and Gordon Moore found himself working as a director of the company’s research and development team. Things were truly great.

Then, after almost 10 years with Fairchild, Gordon Moore and Robert Noyce left to form their own company: Intel Corporation. They had a different vision for how their lab would run. Instead of having research scientists and engineers work separately, they would have the teams work together to develop new chips. This would merge the theory of making circuits with the practice of making chips and deliver an early success with their first magnetic oxide semiconductor memory chip. The rest, as they say, is history.

But, despite his distinguished career in the semiconductor industry, Gordon Moore is best known for an observation he made about the ever-changing semiconductor industry, now called Moore’s Law. In 1965 Gordon Moore was asked to predict developments of the semiconductor over the next decade. He answered with one of the most misquoted answers of all time: The number of transistors per silicon chip will double each year. He later revised this to be every two years, but time has shown us that the number of transistors have doubled approximately every 18 months.

Despite modifying his answer, the last few decades have proven his observation true. The number of transistors we’ve been able to put on an IC has doubled every 18 to 24 months. But, we’re approaching a barrier as these transistors get smaller. This is leading to new challenges that even Gordon Moore could not have predicted. So, please join us again next week as we discuss the future of Moore’s Law and see if it will continue to provide us the increased speed, complexity, and size reduction it always has.

Comments are closed.