AMD Details New Model Numbering System For 2023 Mobile Processors

[Teggs]

AMD seems to be picking up the Intel market segmentation disease.

I'm not sure I agree about segmentation. AMD already used the same segmentation in mobile. But these namings will leave the average consumer confused as hell.

'But I got a 7000 part. You're saying it's Zen 2 and that's old?'

Edit: Never mind. I see that this scheme even comes with a 1980s computer-game style decoding wheel. Screw the segmentation and the naming. This is just rediculous.

[Paradigm Shifter]

I'm not sure I agree about segmentation. AMD already used the same segmentation in mobile. But these namings will leave the average consumer confused as hell.

'But I got a 7000 part. You're saying it's Zen 2 and that's old?'

Edit: Never mind. I see that this scheme even comes with a 1980s computer-game style decoding wheel. Screw the segmentation and the naming. This is just rediculous.

People would understand if they just stuck to a naming scheme. Neither AMD nor Intel seem to be able to come up with a naming scheme and stick with it. I find Intel particularly bad at this - almost to the point of deliberate obfuscation - but AMD chop and change regularly as well.

[Luke_Wolf]

Um... Guys... Guys... Let me let you in on a little secret... It's the same numbering system they had been using but with more information added into it by actually using the last two digits. What actually is the problem here?

[bachchain]

_ - __ ___ __ _
|   |  |   |  |
|   |  |   |   - Optional modifier letter
|   |  |    - 2 digits for max boost freq
|   |   - 3 digits for core count
|    - 1-2 digits for architecture
 - Letter for market segment

There, I've fixed AMD's naming scheme. It can work for up to 99 generations, so it should last until the death of silicon transistors.

[NateHubbard]

3 digits for core count

it should last until the death of silicon transistors.

So my 1024 core processor won't be happening?

[billyswong]

Um... Guys... Guys... Let me let you in on a little secret... It's the same numbering system they had been using but with more information added into it by actually using the last two digits. What actually is the problem here?

The old scheme, without using additional digit for "architecture", let people deduce the architecture from the first digit. (Not totally consistent but usually okay.) The marketing people may like new products to always contain bigger number, but sane buyers will wish old model refreshes be named as proper old model refreshes. Whatever they refreshed or improved outside core architecture should have been minor details that only deserve the "feature isolation" digit at the end.

[stormcrow]

According to the AMD blog post listing out the requirements they were looking for no. 4, making it easier for average joe to know what they're looking at is already a failure. Engineers just do not speak the same language as the general public. Hell, engineers and physicists can't even talk to each other without screwing up signs and terms and arguably they're supposed to be on the same page (engineering being applied physics).

I don't even think many "enthusiasts" will be able to keep up with the scheme unless they're dealing with it extremely often.

This is engineering turned marketing turned... mush. Hint to AMD: You fail as soon as you had the same numbers applying to different SKUs which is how I'm reading the middle two numbers. If you're going to step, then step. Don't use the same number for step 1 as if it's the same as step 2 even if they're using the same basic architecture. (the 7/9 in market segment, gen/gen+ in architecture). That's going to confuse a hell of a lot of people. If you wanted to be more explicit, well... frankly you're clearer than Intel's mud, but that's a ground level bar.

[rabcor]

Man I want a 6800U in a 17" gaming laptop with some dedicated gpu but those things are hard to find, not a lot of models have the 6800U yet (but they really should, it's an incredible cpu for a laptop)

And usually when gaming laptops are made they'll have the H model, maybe HS, rarely U, it's tough having niche needs.

[billyswong]

According to the AMD blog post listing out the requirements they were looking for no. 4, making it easier for average joe to know what they're looking at is already a failure. Engineers just do not speak the same language as the general public. Hell, engineers and physicists can't even talk to each other without screwing up signs and terms and arguably they're supposed to be on the same page (engineering being applied physics).

I don't even think many "enthusiasts" will be able to keep up with the scheme unless they're dealing with it extremely often.

This is engineering turned marketing turned... mush. Hint to AMD: You fail as soon as you had the same numbers applying to different SKUs which is how I'm reading the middle two numbers. If you're going to step, then step. Don't use the same number for step 1 as if it's the same as step 2 even if they're using the same basic architecture. (the 7/9 in market segment, gen/gen+ in architecture). That's going to confuse a hell of a lot of people. If you wanted to be more explicit, well... frankly you're clearer than Intel's mud, but that's a ground level bar.

I think they have already messed up the scheme when they put "2xxx" numbering into Zen1 mobile & APU just because those are released a year late. And now their marketing department still pushes the flawed thinking that consumers will like model numbering that emphasize which year a CPU is first manufactured, which is something no one care and mostly regarded as deceptive marketing.

We can't change existing model numbering but I think my idea is far better than what they came up with:

1st digit (or 1st double digits for distant future): Zen1 => 1, Zen1+ => 2, Zen2 => 3, Zen3 => 4, Zen3+ => 5, Zen4 => 6, and so on. Full step or half step architecture improvement is minor technical detail that don't need to be presented in model numbering.
2nd digit: Here is the usual market segment indication. 1 for cheapest model and 9 for the most expensive.
3rd digit: 0 for original model, >0 for extra number of cores in later refreshes, or for "Threadrippers" in comparison to mainstream models. For example, 1 for 150% core count, 2 for 200% core count, and so on. The exact numbering scheme depends on how many levels of core count increase the company may want to add, and can change across generation.
4th digit: 0 for original model, >0 for lithography, frequency, or power efficiency improvement. Exact numbering scheme here can also change across generation. Buyers only need to be at ease that bigger number in this digit always mean a better product if the earlier digits are all the same.

If the 1st digit or 1st double digits got too big in far future, one can always change the "Ryzen" brand name. The marketing guys will probably want to do that once in a blue moon anyway.

[Sonadow]

Great, now I can focus only on "e" and "U" cpus and forget the rest of the numbers in the name
And hope that some noname chinese vendor picks up these procs and offers nice small quiet passively cooled desktop systems with them. Intel is strongly dominating this segment ...

I know. Those cheap mini PCs with Apollo Lake, Gemini Lake, Jasper Lake and Eikhart Lake processors, socketed SSDs and socketed / soldered RAM are only produced and sold by no-name Chinese OEMs. And they are usually dirt cheap at < $300.

They are so affordable I have at least one mini PC for every generation from Apollo Lake all the way up to Jasper Lake. Running a mix of Linux or Windows 11 based on needs. One of them is my office as my BYOD computer at work.

Naturally, there's nothing from AMD to compete in this segment for the past few years.