May 8, 2020
9
2
15
0
I do hope AMD does skip to 5000-series for Zen 3 to end the generational model number discrepancy between CPUs and APUs.
I wonder if the iGPU is too much of an odd bird in terms of relative performance to the average consumer not even really knowing the difference between storage & RAM. AMD's really re-establishing a name for themselves, maybe they just want to make clear that the APUs are significantly stronger & better without boring the customer to death with explanation of what a graphics card is & how it can fit on the same piece of silicon as a CPU.
 

Gillerer

Distinguished
Sep 23, 2013
315
37
18,890
30
I don't know how the way APUs are one thousand ahead of the desktop parts is till difficult to graps for tech media, when AMD (and Intel for Core HEDT) have been doing it for years.

The APUs come out up to a year after the same-generation CPUs, and the next generation CPUs are always only months away. If they had numbering according to architecture, it would cause more confusion for regular people, when most of the time (8 - 10 months out of the year) the "latest" CPUs and APUs would have different numbers - customers would disregard the AMD laptops because to the uninitiated it would seem they have old processors. It's especially bad that this would occur during the most important sales period of the year - the last quarter.

Obviously tech enthusiasts are annoyed by this, but the majority of PCs are sold to other people.

*

I think there is only one group of people that will potentially get confused: Less tech-savvy people who for some reason think they know about the subject, even without putting much effort into learning.

Everyone who actually knows about this can easily make the distinction.

For everyday Joe, the only thing they want to or need to know is that "4000" is the latest stuff from AMD.
 
Last edited:

InvalidError

Titan
Moderator
I don't know how the way APUs are one thousand ahead of the desktop parts is till difficult to graps for tech media, when AMD (and Intel for Core HEDT) have been doing it for years.
AMD's HEDT uses the same generation numbering as mainstream. APUs are the odd ones out. Also, HEDT uses a different socket and pricing range so there is no chance of actually getting the two confused market-wise. Intel has no such issue in the mainstream since its IGP vs no IGP CPUs are physically the same and only differentiated in model number by the 'F' suffix.

For everyday Joe, the only thing they want to or need to know is that "4000" is the latest stuff from AMD.
To which I respond that the numbering scheme is misleading at best, a potential liability at worst: I could easily imagine someone suing AMD for false and misleading advertisement by slapping next-gen numbers on old-gen tech sharing the same platform.

APUs are one generation behind on tech and model numbering should reflect that.
 
To which I respond that the numbering scheme is misleading at best, a potential liability at worst: I could easily imagine someone suing AMD for false and misleading advertisement by slapping next-gen numbers on old-gen tech sharing the same platform.

APUs are one generation behind on tech and model numbering should reflect that.
I kind of agree with Gillerer that none of this really matters all that much. In the end, the new generation of CPU architecture typically doesn't do a whole lot different from the previous generation as far as the end-user experience is concerned. Maybe the newer architecture is a little faster, but having processors that perform differently within a generation is pretty much expected.

People could similarly hear about how one of the unlocked Intel processors performs at gaming or some other task, then be disappointed to find that the locked part they bought performs 10% slower. In that case, Intel is artificially restricting performance by locking clocks to a lower level than the processor is actually capable of, while in the case of these AMD APUs, performance is limited by not using their newest architecture.

Each generation of processors covers a range of performance levels, and there's a variety of ways for the manufacturer to segregate that range, whether its by locking clock rates and binning chips, disabling cores or SMT, turning off integrated graphics, or using a different architecture. The numbering scheme is pretty much just there to say "This is the new processor lineup for the latter half of 2020".

It might be a little more ideal for these APUs to be using the newest design, though this also frees up manufacturing capacity on the newest process node for other parts, while still letting people know that these are the latest APUs from AMD, and newer models won't likely be coming for another year or so.
 

InvalidError

Titan
Moderator
In the end, the new generation of CPU architecture typically doesn't do a whole lot different from the previous generation as far as the end-user experience is concerned.
From a legal standpoint, the 10-20% differences are more than enough to qualify as materially significant. I wouldn't be surprised if the branding change was caused by such complaints.
 
Again, there can be similar performance differences just by locking clock rates, using a different core configuration, disabling SMT, or any other number of things. And the Zen3 parts are not out yet, nor have any specific products using that architecture even been publicly announced, so it's not like AMD is launching a product that doesn't perform as well with a similar name at a later time. With the G-series processors, one will effectively be getting a processor that trades a little CPU performance for a relatively capable iGPU, for those that have need of it.
 

Flemishdragon

Commendable
Mar 15, 2019
18
0
1,510
0
Sounds already much better than the 4.6Ghz. It's great every spotting it's a bit better. Will it blast the tiger out of the lake? Ryzen Tiger Blaster 4950x.
Ryzen Lava Burster 4980. Vicious nova chill 4990 Because it still runs cool on 64cores.....
 

Flemishdragon

Commendable
Mar 15, 2019
18
0
1,510
0
Having 4.9GHz doesn't mean all cores can hit that speed. Perhaps just 1 core.......
Yes that's a sad reality. But that makes it possible to cool it more easily.
Some people overclocking Intel18,24 on all cores use a sub zero chiller., while that's fun to see or read about on the net. Those are expensive 1grand and more+extra power consumption and also have the risk of condense and such .
 

ASK THE COMMUNITY