The Death of Moore's Law?

In 1965, Gordon Moore, co-founder of Intel, made the observation that the number of transistors per square inch on integrated circuits was doubling every year.

The paper recording these findings was aptly titled, “Cramming More Components onto Integrated Circuits.” It was the basis for what we now know today as “Moore’s Law.”

Moore’s Law predicts that every two years (Moore revised the projection to two years in 1975), integrated circuits would essentially double in power. By cutting the size of transistors in half every 24 months, we’re able to fit more computing power in less space.

In the same 1965 paper, Moore argued, “The future of integrated electronics is the future of electronics itself.” Looking back five decades, it’s quite obvious he was right.

The ever-decreasing size of computer transistors has, so far, been the single-greatest driving force behind advancements in consumer electronics. Smaller chips with greater computing power have put computers, once the size of entire buildings, quite literally in the palms of our hands.

Without the features of Moore’s Law, technological progress would simply not have been possible to the extent that we’ve witnessed over the last 50 years. We would not have mobile phones, tablets, or smart watches. We would not be privy to the connected world as we know it today, either.

MIT Technology Review put it like this back in 2000:

The effect of Moore’s Law on daily life is obvious. It is why today’s $3,000 personal computer will cost $1,500 next year and be obsolete the year after. It is why the children who grew up playing Pong in game arcades have children who grow up playing Quake on the Internet. It is why the word-processing program that fit on two floppy disks a decade ago now fills up half a CD-ROM-in fact, it explains why floppy disks themselves have almost been replaced by CD-ROMs, CD-Rs and CD-RWs.

Yet after five decades of Moore’s Law holding its own, skepticism has begun to emerge. There is worry across the sector that because of physical or economic limitations, Moore’s Law may finally be coming to an end.

If this happens to be true, the implications on the consumer electronics industry could be immense…

Keeping Up With Moore’s Law

In 2013, director of the Microsystems Technology Office at DARPA, Robert Colwell, stated: “My thesis here is that it’s time to start planning for the end of Moore’s Law, and that it’s worth pondering how it will end, not just when.”

Here, Colwell was referring to the idea that Moore’s Law is fighting two distinct battles:

The first is simple physics. For one, as computer chips continue to take on an increasing number of transistors, the heat they produce becomes immense. Very soon, materials such as silicon will need to be replaced with others in order to satisfy these increasingly demanding conditions.

Further, transistors are rapidly approaching the atomic level…

Theoretical physicist Michio Kaku, for instance, believes we have about eight years left of Moore’s Law because of the limitations imposed by the laws of thermodynamics and quantum physics. Once you start talking about building computers at the atomic level, the idea of going smaller obviously gets a bit more complex.

“Around 2020 or soon afterward, Moore’s law will gradually cease to hold true and Silicon Valley may slowly turn into a rust belt unless a replacement [for silicon] is found,” says Kaku. “Transistors will be so small that quantum theory or atomic physics takes over and electrons leak out of the wires. At that point, according to the laws of physics, the quantum theory takes over.”

Now, I’m not going to pretend I understand all the ins and outs of quantum physics (I don’t), but I am willing to put my trust in those who do. It seems reasonable to accept, based on the claims of many physicists, that at some point or another, computers will stop getting smaller — at least at the reduction rate we’ve seen over the last 50 years.

As for the second battle now being faced by Moore’s Law, it’s a simple case of economics.

While it may be theoretically feasible to keep up with Moore’s Law for the foreseeable future, that doesn’t mean it makes sense to do so from a financial standpoint. The smaller transistors become, the more cost intensive it is for companies like Intel and AMD to manufacturer chips.

It’s been argued before that Moore’s Law only works because it’s a self-fulfilling prophecy. That is, because the semiconductor industry uses it as a benchmark for progress, transistor size tends to fall in line with those expectations.

If this is true, it means Moore’s Law is not the product of some inherent rule of science so much as it is a product of the industry’s willingness to invest in the idea of it. In other words, once the benefit of smaller chip size stops outweighing the cost, Moore’s Law may cease to exist.

Post Mooretem

Exactly what the death of Moore’s Law would mean for the consumer electronics industry is not completely clear.

Some would argue that once computers reach the atomic level, there will be little need to reduce size in the first place. After all, we’ve already reached a point where most consumers don’t use their devices to full capability anyway.

For instance, you’d be hard-pressed to find someone who actually needs all 128 GB of storage on their iPhone. You’d also find it difficult to locate many people who utilize more than 8 GB (let alone 32 GB) of RAM on their personal laptop.

Others, though, are keen to point out that hardware manufacturers could suffer greatly once computer chips stop reducing in size. If Moore’s Law means upgrading your laptop or phone every two years for better models, there will be less of a need to replace your electronics once processing power becomes stagnant.

Of course, people don’t just replace their electronics solely to upgrade — sometimes devices just get old.

It’s likely that when Moore’s Law dies, the consumer electronics industry will begin to resemble the automotive industry in a sense. A car built in 1990, for example, is not drastically different in terms of performance than a car built today, but basic wear and tear ensures the need for new models.

In any event, the death of Moore’s Law is likely to have at least some negative impact on consumer electronics companies across the board. People won’t be replacing their mobile devices every 10 years, but they’ll certainly be less inclined to buy an iPhone 10 if it has the same performance specs as the iPhone 9.

Fortunately for companies like Apple, though, there’s still some runway left before we hit a wall. Advances are still being made in the transistor space, with researchers continuously pushing the limit on what’s possible.

Where We Stand

As of last week, a team from Cornell University successfully demonstrated the ability to create transistors just three atoms thick (about 0.6 nm). For reference, Intel’s 7 nm chips, which happen to keep up with Moore’s Law, aren’t even due until 2016.

In other words, there’s a fair amount of time left before you actually need to worry about the end of Moore’s Law. Then again, this doesn’t mean you should be ignoring where the semiconductor industry stands in regards to it today.

The reality is that before we experience the death of Moore’s Law, we will continue to see the semiconductor industry evolve in an effort keep up with it. In the past, this meant little more than developing new fabrication methods. In the coming years, though, it will mean transitioning to entirely new materials.

Intel’s 7 nm chips, for instance, are rumored to use a gallium-indium alloy in replacement of silicon. Cornell University’s “three atom transistors” use a metal called molybdenum.

In the coming years, “tech metals” such as those mentioned above will become increasingly important to the consumer electronics industry, and unlike silicon, these metals are limited in supply.

For perspective, total annual production of silicon is ~900,000 tonnes, while total annual production for molybdenum is currently about 600 tonnes. Obviously this could make “tech metal” exploration and development companies particularly valuable, which is truly excellent news considering that right now, most are trading dirt-cheap.

Until next time,

Jason Stutman

Check us out on YouTube!