“The number of transistors and resistors on a chip doubles every 18 months.”
– Gordon Moore
Gordon Moore, the co-founder of Fairchild Semiconductor and Intel said these words after observing how the development of the computer’s core processor was evolving and this was back in 1965. He then later revised his theory in 1995 and said that the number of transistors and resistors on a chip doubles every 24 months. According to scientists and engineers this law applies to all electronics technology; however, it seems that electronics are still getting stronger and faster year after year. So it begs the question, was Moore wrong? Probably not.
Let’s take all this apart and see if Moore’s Law has become irrelevant or does it still influence the world of consumer electronics.
Moore’s Law Explained
What Gordon Moore was really talking about in his theory was the actual number of transistors that could be fitted onto a core processor microchip. Before we proceed any further, let us first understand what a transistor is. A transistor is a semiconductor device used to amplify or switch electronic signals and electrical power. The laws of physics will always apply in designing CPUs or “central processing units” and you’ll have factors like heat and energy loss that will determine the overall performance of the CPU.
Transistors also consume a lot of power, especially if they are instructed to make tons of calculations, which, again, is a big challenge for CPU designers. Because the goal is to make the CPU as fast and powerful as possible yet does not consume so much power and overheat. For instance, a 10nm (10 nanometers or 1×10-9) chip means that it can store very tiny transistors in it with the spacing between each transistor is 10nm. That’s actually an insanely small size if you think about it and you could fit roughly 3 billion individual transistor cells into it.
You may also want to read Would You Consider the Dell Latitude 5285?
Basically a CPU with all those billions of transistor cells in it is an extremely complicated switchboard or maze with gates that you can give instructions to if you want it to perform specific tasks. The speed in which a CPU operates is measured in “Hertz (Hz)” or 1 cycle per second – meaning it can perform a task at N times per cycle of a second (N meaning any given number based off of the CPU’s computing capacity). So if we were to say that a CPU has a speed of 2.9GHz, then it simply means that it can make calculations at 2.9 billion times per second. Now depending on how big or small or simple or complicated the instructions you give the computer, then it will also determine the length of time it will take for such task/instruction to complete based on its computing power.
Here is where it gets complicated.
Do you remember that switchboard/maze with gates that I mentioned earlier? Well to explain how the CPU works in simple terms, let’s take for instance, that you want to cross from one end of the maze to the other end to get to your destination – in this case you will represent that 5 volts of electricity (the computer’s binary code) that runs through the transistor cells. We have dual, tri, quad, hexa and even an octa-core processor now since the inception of electronic computer devices, which means that a CPU labeled 3.6GHz, but has a quad-core design will work 4 times as much as a single core processor! So going back to you wanting to cross the maze, if it is built on an octa-core design, then you will traverse the gates at 3.7GHz multiplied by 8 times!
It’s crazy fast, but unfortunately, there is a misconception that the CPU will only become more and more powerful in exponential terms as time goes by, which Moore’s Law contradicts. And, in fact, Moore’s Law has already won the debate ever since it was conceived just to be clear. Even now with octa-cores there are only around 0.3 – 0.5GHz difference between CPU brands and that’s because the nanometer design is reaching its limits. Although companies like Snapdragon, Samsung and Apple are still planning to make 7nm, 5nm and even 3nm chips in the near future.
Engineers also said that while they could fit more transistor cells into a CPU they are likely to operate at lesser capacity than one would expect as heating problems will destroy the device even before its warranty expires. They also noted that as transistor cells get smaller and smaller, their capacity will remain constant – meaning they will not get even more powerful like most people would expect.
So What Happens Next?
It is obvious that the demand for a more powerful computer chip and even those that eventually will become A.I.s (artificial intelligence) is growing on a daily basis. So how would computer manufacturers satisfy these demands if Moore’s Law were to reach its limit and bar the CPU from performing any faster? Moore’s Law predicts doubling, but when computers go from quartz to quantum, the factor will be off the scale. As you may already know the current computer technology operates by using semiconductors, metals and electricity. Future computers might use atoms, DNA or light.
The future compact computer could be as small as a dime, but will be a thousand times more powerful than the most powerful computers of this age. So as soon as Moore’s Law bars current computer technology to go beyond its limits, then it is very possible that scientists will have to redesign the CPU at a quantum scale to make a new kind of computer device.