Looking Back at Moore’s Law

This week marks the 50th anniversary of Moore’s Law. We wanted to mark this occasion by taking a look back at this famous prediction and what it’s meant to the computer industry.

Driven by the pursuit of engineering excellence, we now make more transistors every year than there are stars in the Milky Way. Some 25 times more, actually. A number like 250 × 1018 is quite literally incomprehensible but that’s how many transistors were produced in 2014. That’s what enables our technological world.

Let’s take a look back to where it all came from.

computer-chip-1

What Is Moore’s Law?

Moore’s Law is easy to summarize: The number of transistors that can be placed on an integrated circuit will double every eighteen months. That’s the “law.”

You’ll notice right away that it’s an observation, not some natural law. Keep this in mind.

How did Moore get to this observation?

On 19 April 1965, Moore published an article in the trade magazine Electronics, “Cramming More Components onto Integrated Circuits.” He’d been asked to summarize the state of the electronics industry, so he wrote about the potential of a relatively new product: integrated circuits.

In 1961, the company Moore worked for, Fairchild Semiconductor, had made the first planar integrated circuit. It had 4 transistors on it. By 1965, they were up to 64. 4, 8, 16, 32, 64… this basic pattern is what Moore noticed. He simply extended it for another ten years.

Here’s what Moore actually wrote in his short, dry and legendary article.

In section four, “Costs and Curves,” he writes: “The complexity for minimum component costs has increased at a rate of roughly a factor of two per year… That means by 1975, the number of components per integrated circuit for minimum cost will be 65 000.”

Exciting, no?

He would modify his predication in 1975 to “per two years,” but there you have it. That little guess based on what Moore was observing in the integrated circuit world in 1965 is the thing itself.

(It would later get further refined to every 18 months or so, not two years.)

It’s worth backing up for a second to give a bit of context. Moore, who would co-found Intel in 1968, was working as research director at Fairchild Semiconductor at this point. He and seven other engineers and scientists had formed Fairchild in 1957 to make useful, high-speed transistors.

We won’t get into the whole back story, though if you’re interested I highly recommend Michael Riordan’s great article on Jean Hoerni, the man who invented the planar integrated circuit. Riordan does a brilliant job of giving the history.

What’s important to take out of all this is that Moore was focused not exclusively on the sheer engineering capability, but on making integrated circuits useful and economical.

Which brings us to our question:

computer-chip-2

Why Is Moore’s Law Worth Talking About?

Moore’s Law is the standard for progress in technology. This is its fundamental importance: it’s a statement of intent.

As Roger Cheng put it in his retrospective for CNET: “Moore’s Law is more than a guideline for computer processor, or chip, manufacturing. It’s become a shorthand definition for innovation at regular intervals, and has become a self-fulfilling prophecy driving the tech industry.”

Moore started off his article with a bold claim: “The future of integrated electronics is the future of electronics itself.” And he was right. As integrated circuits have gone, so have electronics gone. Because until only recently, the smaller the transistors got, the more efficient and less expensive they got—an almost unique win-win-win situation—electronics have been carried along on the tidal wave of innovation.

Which gets us to our second point.

Moore’s Law is important as a prophecy that’s largely come true, and frankly we humans love a good prophecy. In that article, he wrote, “Integrated circuits will lead to such wonders as home computers… automatic controls for automobiles, and personal portable communications equipment.”

Sound familiar?

Remember that the first home computer was still a decade-and-a-half away at this point. We’re only just now getting self-driving cars. But Moore could envision how these simple on-off switches had the potential to enable wonders.

The biggest reason Moore’s Law is worth revisiting right now is the economic one. It’s interesting to revisit Moore’s original article, because in a way he’s emphasizing not so much what is possible, but what is economically feasible.

He makes a prediction of how many components can be integrated “for minimum cost.” He writes in section eight, “The total cost of making a particular system function must be minimized.”

Charles King writes on Pund-IT: “Over time, Moore’s thoughts on the exponential improvement of technology evolved from a simple observation into a promise that businesses, investors and consumers continue to bank on.”

What Is the Future of Moore’s Law?

We are now at the point where Moore’s Law could stop being effective. Actually, the cycle has already slowed down. Smarter minds than mine are speculating about where the transistor business will go next.

We’re going to leave this topic for now. Next week, we’ll talk about some of the amazing materials and methods that people are coming up with to continue the innovative path that Moore predicted.

Until then!