Let’s Talk: AMD’s Laptop APU Naming Schemes

Let’s Talk: AMD’s Laptop APU Naming Schemes

Let me get this out of a way: AMD hasn’t exactly have the best track record when it comes to giving its product names. It has happened for many years, but has largely been more ‘sensible’ in the Ryzen era. Then Ryzen 7000 mobile happened – with people arguing that the new naming system is at best, too difficult to grasp; or at worst, intentionally misleading. Whichever answer you get depends on person in whatever corner of the Internet you ask, so let’s talk about it.

AMD CEO Dr. Lisa Su revealed the Ryzen 7000 series APUs during CES 2023. Image: AMD (YouTube)

History: It Depends On The Competition

If you take a look at all of the past naming system in AMD’s products, you may have seen a line of CPUs that was particularly interesting when it comes to names. Remember the old Athlon? (Not the new ones that got relegated as entry-level lineup.) The youngins among you may not know, but early 2000s was basically AMD’s finest hour. It is largely speculated by the public at the time that, in order to compete with Intel’s massively popular Pentium lineup of chips, AMD decided to name its CPU against Intel’s clockspeeds with the “PR Rating System”.

(Jump to 3:06 for the video explanation.)

For example: take Intel Pentium 4 2.0GHz – as it says in the name, this is a chip rated for 2.0GHz. Simple. But we all know by know that clockspeeds aren’t the be-all-and-end-all when it comes to CPU specs, because at that time, AMD’s contemporary Athlon XP series was beating Intel equivalents in performance despite having markedly slower clockspeeds. (The basic explanation is down to Instructions Per Clock, or IPC. AMD had a much better IPC due to its core design as opposed to Intel’s.)

The competitor for that Intel chip is AMD Athlon XP 2000+, but guess what? The specs sheet says the clockspeeds are only 1667MHz (~1.67GHz). During those times, the general public wasn’t very well informed of the concept of IPC, hence to market these objectively more performant CPUs, AMD decided to name it against its Intel equivalents and append a ‘+’ sign to signify the fact that they’re the faster chip.

This is just one example – but if you look at how Ryzen desktop processors was named, similar principles apply. Ryzen 3, Ryzen 5, Ryzen 7, or even Ryzen 9 – makes it very easy for people to know what they’re pitting its models against. Arguably, even their GPU division does the same (the odd jump from RX 500 series to RX 5000 series was believed to be matching NVIDIA’s 4-digit naming system as theirs had reached its 10th iteration of modern GeForce naming schemes). And for the most part, they kind of worked.

Ryzen 7000 Mobile: AMD’s Rationale

But AMD decided to do away with the fairly ‘straightforward’ naming scheme that they had been using since 2017 which largely mirrors that of Intel’s, since the Ryzen 7000 mobile series. (Desktop SKUs are so far unaffected.) AMD cited a few reasons, but here’s a few important ones (from this blog post here):

• Our current naming system for Ryzen Mobile Processors was at an end. It simply could not accommodate the influx of new SOCs in new categories we’re developing.
• We wanted the system to be technical and informative, such that an enthusiast could quickly “decode” the number to see what’s inside.
• We wanted it to be simple for the average user, where a higher number simply connotes higher CPU performance.

Now, Intel and AMD – they have somewhat different targets when it comes to the laptop market, and so does their marketing strategy. Intel’s naming doesn’t produce a whole lot of definitive information aside from two things: CPU generation (e.g. 13xxx) and suffixed letters as the CPU class (e.g. H, U, P, HK). While AMD had very similar naming beforehand, this time around they chose to laid out all information upfront with what kind of CPU it’s trying to sell.

Take Ryzen 9 7945HX. Now, if you’re coming from the previous generation – this may get you lost for a bit as AMD’s previous mobile flagship is named 6980HX. 80 vs 45, how does that work? Well, it turns out the old naming system is more of putting an arbitrary number in relation to other models. So in theory, the closer to the hypothetical ‘6999’ value, the better the chip is. But that presents a problem – the number itself doesn’t exactly say what it does.

AMD produced a chart to explain what each digit means. Image: AMD

So the 7945HX basically tells you these few things: 7 is what AMD calls ‘portfolio year’, so anything coming out of this year regardless of which design will automatically starts its model number with 7. The second number, 9, more or less follows previous generation’s naming: the higher the number, the better.

Third number is the most important of all – this tells you which Zen architecture it’s used. If you want the latest and greatest, the number would be ‘4’. 3 stands for Zen 3, 2 for Zen 2, and so on. And finally the number ‘5’ is more or less a ‘separator’ to differentiate between two nearly identical parts on the same gen. Finally, the suffixed letters stays as-is: HX for high-power, HS for low-power and U for ultra-low power, and so on.

Opinion: Has It Worked?

Even though in a perfect world, AMD’s naming system would make sense and is even more informative than what the Intel chip says; but in reality, that doesn’t seem to be the case. People complained about its confusing nature due to everything starts with the number ‘7’ – and they’re probably right. Here’s the thing: humans are wired to read the first letter or number in a sentence, name, or a phrase.

When AMD decided to move the most important number to the third (or even fourth) digit, it inherently creates confusion to consumers even though the name is technically more accurate. Think about it:

7735U vs 7730U
or;
6800U vs 5800U
(Note: 7735U and 6800U are technically not the same chip, they’re just highly similar – but for the sake of this comparison, let’s assume they’re the same. Same applies to the other two chips.)

which one is easier to tell apart? Exactly. AMD has seemingly borrowed its naming system from their datacenter EPYC lineup, which has a completely different group of customer base and the tech literacy is literally worlds apart. So while I appreciate AMD’s efforts in making this more clearer, I think a better naming system is still needed to keep it relatively simple for the general public while concise enough for us, the tech nerds.

But hey, at least they’re no monitor manufacturers who loves to mix a bunch of alphabets and numbers in seemingly random order.

Info source: AMD | Videocardz