How To Pick a Laptop?

Or how to navigate confusing numbers

That was it. The battery died after 10 minutes of use only. The screen split after one too many trips to the library. The keyboard never recovered from its unexpected early morning coffee shower…it’s time for a new laptop.

You open your favorite tech store website, and you start browsing.

Some things, you understand. For example, the size of the screen. 14’’, 15’’, etc.A bigger screen means a higher price (everything else being equal. Which, let’s be honest, it never is) and a heavier weight. The choice is pretty straightforward and depends mostly on how bulky of a machine you can accept.

Then some more numbers…sometimes not as clear. For example, the “RAM”. 8, 16, 32 GB. Here you have to pick 16GB or higher. Every. single. bit of modern technology is hungry for this precious RAM. so, 16GB it is. Sometimes you see “DDR whatever”. Don’t worry about it.

Hard drive! 512GB, 1TB, etc. I think people are generally aware enough of this metric nowadays. The only gotcha is “HDD” (hard drive disk) vs “SDD” (solid state disk). Pick SDD.

And, finally, the utmost confusing: processors…as your future computer’s brain, this is maybe the most importance choice. Yet, you will probably glaze over this alphanumerical imbroglio before ignoring it and just “trust your guts”. Let’s try to bring some clarity.

First, there are two main brands of processors: Intel vs AMD. It doesn’t matter. Pick whichever. If you know you are using a lot of applications at the same time (and I a mean a lot of heavy applications, not just notepad and firefox), then you may prefer AMD, but it’s marginal. Next, how does one compare between models of the same brand?

Before diving into that, let’s get some keywords out of the way:

  • Processors have what is called a “generation“. It’s a fancy term to say a new version. Just like you have an iPhone 16, then 17, etc. Sometimes, there are big changes between versions, sometimes not.
  • Then, another keyword is “tier“. That’s like the model; for example, the iPhone 17 “Pro”.
  • And finally, we have the “performance level“. The same model can be optimized for different energy consumption. Some will be more powerful (but will eat more power) and others will be optimized for less intense, longer sessions.

Knowing that, it’s important to remember that a higher generation does not necessarily mean better. It’s like comparing an iPhone 17s and an iPhone 16 Pro. You can see how the Pro may still be a better choice, even if the version is lower. Same for processors.

Okay, so what does it look like concretely? Let’s start with AMD Ryzen processors. This article will be outdated awfully quick, but, in March 2026, the latest generation is called “AI 400”. Yes, they used the word AI and yes, it’s mostly marketing (see the “NPU” section below). So, for example, you will see Ryzen 423, Ryzen 456, etc. Here is the list of generations:

  • Ryzen AI 400
  • Ryzen AI 300
  • Ryzen 8000
  • Ryzen 7000
  • Ryzen 200 (notice the lack of AI buzzword here)

No, it’s not you, it’s them. There are no ryhmes nor reasons, it’s just to confuse people. But now you know!

Alright, that’s already a big deal. You can now distinguish between the processor equivalent of an iPhone 14 and and iPhone 17. What about models? There are only 3,and those are numbers:

  • 5 is entry level (ie the iPhone S)
  • 7 is mid range
  • 9 is high tier (the Pro)

And finally, the performance optimization. HS and HX means power hungry and powerful. U and e means optimized for less intense tasks.

There you go! You know all of it! Let’s practice with actual processor names I just found on a real website:

  • AMD Ryzen AI 7 350: it’s an “AI 300” generation, so quite recent. It’s a tier 7, so mid range. And they did not give its performance level.
  • AMD Ryzen 9 270: it’s a “Ryzen 200” generation (super duper old!!). Tier 9, so that’s good, I guess…for such an old model.
  • AMD Ryzen 9 8940HX: It’s a 8000 model! Not the newest, but still okay. It’s tier 9, so pretty good. HX performance, so geared towards heavy duty tasks!

See? It’s kind of fun!

Alright, let’s do Intel now. Here is the generation list for them (again, as of Today, March 2026):

  • Core Ultra
  • 14th Generation, today sometimes just called “Core 5” or “Core 7” (notice there is no “Ultra”)
  • 13th Generation, today sometimes just called “Core 3”
  • 12th Generation

As you can see, Intel is even trickier than AMD. The tiers are the same: 5, 7 and 9. The performane level are: H for high performance and U for ultra low power.

Practice time!

  • Intel Core i9-14900HX: This is 14th generation, tiers 9, high performance!
  • Intel Core i5-13420H: this is 13th generation (so it’s quite old now), tiers 5, high-ish performance!
  • Intel Core Ultra 9 275HX: This is the Core Ultra generation (the latest one!), tiers 9 (the highest one!), high performance! This is the best as of today!

Finally, since this article will be outdated before it is published, here is a website that ranks most (if not all) processors. If you ever have a doubt, check it out: Passmark Software - Laptop Leaderboard

And that’s it! You now are able to read all those weird tags and tell what is a good deal and what is not. Congratulations!

Wait…did someone say NPU? Alright, let’s see…ignore it! Bye!

Okay fine, here are a bit more details. You probably heard about some fancy chat bots called ChatGPT, Claude, Gemini, etc. They are very cool, but they use some special math that is very heavy on the processor. Seeing this, hardware manufacturers thought it would be cool to have a special type of chip whose sole job would be to handle those special math. Good idea in theory. However, in practice, it’s up to the software developers to explicitly make use of the chip. As I am writign those lines, the number of app using this is almost exactly 0. The same example is given everywhere: blurring your background on a video call. There. If you want a more efficient blur, that’s your chip. Possibly, in the future, more applications could use it. If the computer you buy comes with it, all the better. It’s very marginal though, so don’t worry about it.

And there we have it! Now, it’s the end of this ar….what? What do you mean “GPUs”? Okay, well…those are kind of big deal. Assuming you play a lot of video games, make those video game, or process huge 3d scene rendering. I am simplifying a lot here. The truth is, everything I said above still stands. But GPU are a beast in and of itself, so that will have to wait for another article.

And now, I am ending this article before anybody has any more ideas! Bye!

This article was updated on