Moore's Law: The Reason Why Your Laptop is Already Obsolete

No matter what kind electronic device you're reading this on, a better version of it exists or is well on its way. If that makes you angry, blame a physicist.

Physics is tricky business. Other areas of science (chemistry and biology for example) can get fairy complicated but there is a certain elegant logic behind them that allows the average person to grasp technical concepts after they put in a little effort. Physics is different. Try as you might to apply overarching logic to the big problems in physics, you will come up short. Even the greatest minds that history has yet produced are unable to explain why you need separate sets of “laws” to describe things that are normal sized or large (baseballs, planets, stars, etc.) versus things that are very tiny to incomprehensibly small in size (electrons, quarks, neutrinos, etc.)

But as dense as physics can get, it isn’t something that we can afford to get frustrated with and give up on. Like it or not, we have built a world in which the seemingly trivial problems of physics (ex. you can either know the location or the speed of an electron, but you can’t know both) actually matter. Case in point is the concept known as Moore’s Law.

Moore’s Law isn’t actually a law, it’s an observation first made by Intel co-founder Gordon Moore in 1965. Put very simply, it is the notion that computing power (the ability of your laptop to do things for you) doubles every 18 months. Put somewhat more technically, it states that the number of transistors on a one inch piece of silicon doubles every 18 months. These descriptions are seemingly similar, but where the former allows for some amusing comparisons, the latter will make your head spin with it’s details.

Let’s start with a fun comparison. Since computing power is said to double every 18 months , it is important to know what that means. Doubling is an exponential rate of growth. In other words, if a laptop in a given year is said to have Sketchy Science Power Rating (SSPR) of 1, the next generation would have an SSPR of 2 (doubled), the next would be 4, then 8, then 16, then 32. It isn’t hard to see that things will quickly get out of hand. In only 15 generations, our computer will be humming along with an SSPR of 32,768. Translated into real examples, Moore’s Law explains why the chip in your birthday card that sings you a song has more computing power than the Allied Forces in WWII; or why your cell phone is more powerful than NASA was at the time of the first moon landing.

To explain how this is possible, we need to understand a little about how electronics work. We are all familiar with the concept of microchips (or microprocessors, if you prefer). They are the brains of our technology. Many things help determine the speed of a microprocessor but one of the most important is the number of transistors on it. Transistors are like switches that control the flow of electrons, and controlling electrons is how you make them do work for you, hence the label “electronics.”

Transistors are the modern substitute for vacuum tubes (those big lightbulb-like things you see on contraptions in mad scientist movies from the 1940’s). The first transistor was built in 1947 and was 1.3 cm (half an inch) long. Since then, to put things mildly, we have gotten really, really, ridiculously good at making them. Modern transistors are about 40 nanometers in size. For perspective, a human hair is 100,000 nanometers thick. As you can imagine, we can now cram a lot of transistors onto a microchip. In 1965 when Gordon Moore made his observation, he used it to predict that in ten years the number of components on a circuit (read: transistors on a chip) would grow from 50 in 1965 to around 65,000 in 1975. He was correct. The trend has since continued and Intel’s current i7 microprocessor now contains 731 million transistors and it’s Xeon processor had 1.9 billion. No wonder new laptops are obsolete every few months.

This is all very exciting, but why exactly should you care? Well, the answer lies in the fact that eventually Moore’s Law will cease to be true. The problem with building things on the scale of nanometers is that you become subject to the laws of quantum physics. When you can’t know both the location and speed of an electron, it’s becomes very hard to control them. When transistors get too small electrons can start to tunnel through seemingly solid materials and cause chips to malfunction.

World renowned physicist and ballsy future predictor, Dr. Michio Kaku has argued that Moore’s Law will reach an end sometime in the next 10 years. To be fair, the great minds of physics also predicted an end to Moore’s Law in the 60’s and the 80’s, but their predictions were based more on humans running out of ideas about how to organize things on a chip and less on the actual laws of physics.

When Moore’s Law ends we will need to transition to new forms of technology if we want things to continue to gain power. Some have predicted computers based on the structure of DNA, or ones that actually operate on the quantum level (quantum computers) that may only be made up of a few atoms. Regardless of what the new technology looks like, the transition will likely be a bumpy one. The global economy will suffer and the price of electronics might take a roller coaster ride for a few years. It’s a challenge that the physicists are already working on. The least we can do is try not to gloss over or call them nerds when they go from sounding almost understandable into completely insane.

Good luck physicists. Now where is my iPod?