What is a CPU? A Simple Guide to Processors, and How to Tell the Difference

If you've spent any amount of time trying to get to grips with computers and how they work, or you want to build or buy one for yourself, you'll almost definitely have heard everyone in the technology space referring to a CPU. But what is it and what does it do, and why is it important to understand?

What is a CPU and What Does it Do?

Let's start this guide off simple - what is the pupose of a CPU?

The CPU or processor is the core to any system, as it is the main unit that receives instructions and distributes tasks to other parts of the system. Everything your computer does goes through the CPU, so it's vital that you make the right choice.

If you're noticing your computer slowing down, there's a good chance that your CPU could be one of the causes - another thing that will point to this is if (particularly in a laptop) your system gets loud and hot in normal, everyday use, as your CPU may be being stressed to carry out what you're trying to do on your PC

How easy is it to upgrade a CPU?

This depends on what type of system you have. On a desktop PC, they're designed to be easy to replace. Usually, you just need to remove a cooler, and then the CPU can be unclipped and removed, and replaced with a different one. If you're working with a very small form factor system (like an Intel NUC) or a laptop, there's a good chance the CPU is not upgradeable as it is very likely attached directly to the motherboard and next to impossible to remove or replace.

One important thing to bear in mind when upgrading your CPU is motherboard compatibility - if you're upgrading to something more than a couple of years newer than what you currently have, you'll often need to replace the motherboard too. More on that later.

CPU Nomenclature Explained

What I mean by this is the combination of letters and numbers that make up the name of any commercially available CPU.

Let's start with Intel!

Intel have a pretty consistent style when it comes to the naming conventions of their CPUs. Since roughly 2009, Intel have been using the same "Core i" naming scheme which now refers to four different levels of mainstream processors - i3, i5, i7 and i9.

I will use the latest generation as an example of how they generally look:

i9-13900K

The green section denotes the product tier. Starting with i3, considered "entry level" the performance steps up with each line from there. i3 are the most basic, usually found in home or office PCs, i5 are "mid-range" and are usually in business or gaming computers, which require good performance, most of the time. i7 is "high-end" and is commonly found in workstation PCs that require strong performance while, for example, working with digital media, music, or gaming at high resolutions. Finally, i9 is the "enthusiast" tier, which is for the most intensive use cases, normally gaming, high-resolution video editing, professional photo editing and the like.

The blue section is the generation indicator. The current Intel processors are in their 13th generation, and previous generations will have different numbers respectively. Before the 10th generation, there would only have been one number to indicate generation, for example i7-6700. Only the first number after the product tier indicates the generation, in this case, 6th Generation. First generation Intel "Core i" CPUs had no generation indicator, for example the i5-650. If you see an Intel processor with only three numbers in its name like this, you should probably avoid it, as they are very outdated now.

The pink section denotes the product SKU. There's not much consistency to how this part works between generations, besides that it will always contain three numbers. You don't need to pay too much attention to this really, just keep in mind that higher is better, and a higher number will be reflected in a higher price. For example, with 13th Generation i5 processors, you will see many different SKUs, and the i5-13500 has more cores and runs slightly faster than the lower-priced i5-13400, despite being in the same product tier. These differences are often not large enough to make a great deal of difference in use, but it's something to keep in mind if you can find a CPU with a higher SKU discounted, for example.

The yellow section is the product suffix. There are a few key things that the letter at the end of the name will denote, which I've listed simply below:

- K = unlocked, meaning you can tune voltage and clock speeds to optimise performance.

- T, U, Y = low power, optimised for mobile. This usually also means lower performance, as the wattage is restricted. These could be a good choice if you want a small form factor system which would have limited cooling capacity.

- F = no on-board graphics. This is an important one to take note of, as your system will require a dedicated graphics processor to display an image if you have one of these.

There are more variations than this, but generally you will see one of these, or a combination of two of these together, like KF, which would mean unlocked and no on-board graphics.

AMD Nomenclature

AMD processors have changed their naming scheme a couple of times, but as a general rule, if you see one that is something like Phenom ii or Athlon ii, avoid it. Likewise, avoid AMD processors that start with an FX. They're just too old and not up to the task any more.

What you want is a processor with the Ryzen model name. Since Ryzen was first introduced, they have had a fairly similar naming convention to Intel, so thankfully once you understand the basics it's not too hard to get your head around.

I'll use the current generation as an example:

Ryzen 7 7800X

Green indicates the product tier. Similarly to Intel, the main product tiers go from Ryzen 3 to 5, 7 and 9, 3 being considered "entry level" and stepping up with every tier from there.

Blue denotes the generation. This is very similar to what was already explained on the Intel side but unlike Intel, the 1st generation products did have a 1 in the name, for example Ryzen 5 1600, which then became 2600 for the next generation. Hopefully that makes some sense!

Pink indicates the product SKU. I feel like I'm saying this a lot, but yes this is almost exactly the same as how Intel processors work. The higher the number, the better and it indicates variations within the same product tier. For example, a Ryzen 5 1400 has four cores while a Ryzen 5 1600 has 6 cores, but the difference in performance is not so extreme that it should be moved into a different tier. Capiche?

Finally, yellow is the product suffix. Once again, this is very similar to Intel processors but there are a few less suffixes to look out for overall, and/or they just make more sense, which is nice. The most common and important ones I've detailed below:

- X = technically means unlocked, but this is less important than on Intel CPUs because all Ryzens are actually unlocked to an extent. X means higher TDP (Thermal Design Power), which means it can be overclocked further and draw more wattage. Better for enthusiasts.

- X3D = the gameriest of all gamer chips. These utilise a 3D V-Cache which, let's just say is a tremendous boost in games.

- G = G is for Graphics. AMD CPUs differ from Intel in that some have very strong on-board graphics. These are referred to as "APU"s and will always have a G at the end of their name.

- U, H = optimised for mobile use, e.g. laptops and handhelds. Lower power, lower performance.

There's not much more to it than that. When you understand one, you can easily make sense of the other.

CPU Jargon Translator

Here, I'm going to break down some commonly used jargon to try to make it all make a bit more sense.

Cores

Cores are as they sound - they are the part of the CPU that data is processed through. How fast it can process that data is dependent on a few things, but particularly how many cores it has. Generally, most desktop CPUs these days have at least four cores.

Applications like video encoding need lots of cores, and the difference in performance tends to scale well with the amount of cores in the system.

In games, once you get above 4-6 cores, the amount of cores is not necessarily as important as the performance of said cores when it comes to how well a game will run. Games need those cores to be fast - run at high clock speeds - (which is why Intel processors are often the fastest in games, despite having fewer cores) and need to have strong multi-threaded performance, which is often where big improvements are made generationally. If that doesn't make sense, read on to the next section!

Single Threaded vs Multi Threaded

You're going to see these phrases in almost any specific CPU review you look at. To boil it down to its most simple form, single threaded applications are those that only run one task at a time, like a calculator or text editor, like spreadsheets or Word documents.

Multi threaded applications make up the majority of modern programs and games, and those are applications that are able to utilise multiple cores at one time to run tasks simultaneously. Video encoding is a good example of a heavily multi-threaded application - it will use every core available to it. To an extent, almost everything is some kind of combination of both, so when you see the benchmarks for single and multi threaded performance it's mostly useful as a comparison for performance between different processors.

If you're building a PC for gaming, keep in mind that most new games need to be able to run multi-threaded applications well, which is why there has been a drastic shift upwards in the amount of cores you'll see in consumer processors today.

Clock Speed

If you're someone who was around computers in the early 2000s, you'll know that for a time, clock speed was everything, and the speed of the processor was often advertised on the front of the computer - weird! Clock speed is usually expressed in GHz or Gigahertz and, less commonly, MHz or Megahertz, and it's an indicator instructions can be processed per second. It's relatively important when comparing performance but much less so now than it was a decade or more ago.

Intel CPUs generally favour high clock speeds, which is often combined with high power usage, and AMD CPUs tend to favour more cores over clock speed for exracting performance, but in the end they perform similarly, often to within a few percent of each other.

What about Laptop CPUs?

Admittedly, this post has been very heavily focused on desktop CPUs, and I have let mobile processors slip to the wayside. Mostly, this is because they are often designed on different architectures. Desktop CPUs are designed to favour performance with less regard to cooling or efficiency, as cooling solutions for desktops are very flexible and interchangeable. However, here are some key points about mobiles:

• In smartphones, you will usually see a Qualcomm Snapdragon in Android devices, or less commonly Exynos (in Samsung phones in the UK), or MediaTek. Broadly speaking, these are all roughly equivalent to each other at different price tiers, but there is so much variation in the market as a whole it would need an entire blog post dedicated to just that.

• Apple devices will use the Bionic processor, which Apple makes in-house. These tend to be significantly more powerful than alternatives and are extremely well optimised for efficiency and things like memory usage.

• Google devices have recently started using Google's own in-house brand, Tensor. Similarly to Apple, these are optimised specifically for Google devices so you will get buttery smooth performance and impressive AI performance with Tensor.

Intel and AMD are still the main players for laptops and handhelds, but expect the landscape to shift slightly in the coming years. ARM processors (very broadly meaning CPU integrated into the motherboard, like the above ones found in smartphones) have made huge gains in the last few years in performance and efficiency, to the point where a top-of-the-range Samsung or Apple smartphone is as powerful, if not more powerful, than most consumer PCs.

Such power has prompted Apple to start making their own ARM processor, imaginatively referred to as Apple Silicon, starting with the M1 processors and now moving on to M2. In MacBooks, these were so unbelievably fast and efficient, they pretty much obsoleted their previous-generation Intel counterparts overnight, sipping power and running silently whilst delivering almost three times the power - and yes, you read that right.

I would certainly expect other manufacturers to be taking note, and unless Intel and AMD make great strides in mobile processors very soon, there may well be a lot more players on the market for laptops, handheld and mobile devices.

Take Your Knowledge to the Market!

This article has been a simplification of CPU definitions and jargon, which I hope will empower you to take greater confidence into buying your next computer, or to upgrading your current one. I have another article if you need help finding the best CPU for your gaming needs, but the key thing to remember if you're getting into this from the ground floor is that, at any particular price point, performance between manufacturers is very similar, and which one is the right one for you should be down to your preference in terms of your usage, price, efficiency and so on. There's no right or wrong answer.

It also comes down to compatibility - if you're simply upgrading, have a look at what your current motherboard is compatible with, as that will help narrow down your search field significantly.

Most importantly, read reviews. You won't know how good something is relative to what you have or relative to its competition simply by reading the description. There's loads of outlets from which you can find comparative tests, so it's never been easier!

Leave a comment if you have anything else to add, or head to my website and send an enquiry if you would like some help assembling your perfect PC - I offer a building service!