Has CPU Performance Reached a Plateau?

Hello guys,

Today I thought a cool topic to write about would be CPU performance and whether it has reached a plateau.

_[1]

You probably have heard of Moore's law which states that computer power doubles about every two years, this law actually is talking about transistor density because in CPUs, the actual power of the CPU is pretty much determined by the number of transistors. That's why you can see really old CPUs that have the same frequency like 3.0 gigahertz but there are a lot weaker than today's CPUs obviously because they have way fewer transistors.

_{[2]
[Exponential growth with transistor count doubling every two years Wikipedia]}

The CEO of Intel said that at this point, computer power really only doubles about once every 2 years so that's kind of a cadence we are in right now.

How does computer power increase?

As I said before, it's all about transistors in a CPU. Basically, as the transistor get smaller, you can fit more in and do more stuff in the same amount of space but obviously if it get super small, there's going to be a limit to how small you can make a transistor.

We are currently about 14 nanometer architecture and they're probably going to push it down to 10 pretty soon but when you get small enough, you’re not going to be able to get any smaller because there is a fundamental limit which is the size of an atom and even bigger than that, once you get really small down to the size of a couple atoms, you're going to get some atomic forces that are going to interfere with how the transistor would be able to operate.

_{[3]
[Intel 14nm technology in detail Anandtech]}

Transistor density is only one option for increasing computer power obviously, you can't get any smaller than a certain point but you can increase the actual size of the CPU itself using the same density so you gotta add more transistors, it would just have to be bigger. We could also see higher gigahertz and higher frequencies to improve the processing speed, the number of operations per second so you could have the same number of transistors but make it more powerful if we figure out how to increase the frequency like the crazy new technologies altogether such as quantum computing.

A transistor works on “1s” and “0s” on binary but quantum transistors or computers actually has three, it's called a qubit, it can be either on = 1, off = 0 or both at the same time so I don't really know the specifics of that but that would be able to exponentially increase the power of a computer.

As a side note, there are actually some developed quantum computers that do work and are able to calculate very basic operations.

For five decades, Moore’s law held up pretty well: Roughly every two years, the number of transistors one could fit on a chip doubled, all while costs steadily declined. Today, however, transistors and other electronic components are so small they’re beginning to bump up against fundamental physical limits on their size. Moore’s law has reached its end, and it’s going to take something different to meet the need for computing that is ever faster, cheaper and more efficient.
_source

Moore's law has pretty much been true for the past several decades but I'd imagine we're going to start to see it not hold true for a while, it might slow down but obviously, computer power is never going to stop improving as technology moves forward, we just might not be able to get past that fundamental transistor density and that’s simply because of the atoms, you can't get smaller than an atom.

Conclusion

As I mentioned, there's going to be other stuff we can do to improve efficiency even in the software side so I'm not too worried and I think we're going to start to see some amazing technology computers coming in each generation.
I'm sure you guys have some thoughts on this so share your thoughts and don’t forget to smash that upvote button.

---

---

References: 1 - 2 - 3

Image sources: 1 - 2 - 3

---

---

_{GIF made by @foundation}

---

---

Thanks to @camzy & @samstickkz for this amazing toon.