Intel Rebranding "Core i" Brand to "Core Ultra"

[Sauron]

thanks I hate it

[HenrySalayne]

10 hours ago, Arika S said:

OK, but why?

 

So Intel is never allowed to rebrand their stuff ever?

 

Was amd more or less pathetic when rebranding from athlon fx to fx and then fx to ryzen?

Let's see:

- Ryzen - trademarkable

- Athlon - trademarkable

- Pentium - trademarkable

- Celeron - trademarkable

even:

- Core - trademarkable

- FX - trademarkable

 

But:

- Ultra - non-trademarkable (because it's already in use by many other brands)

 

 

AMD and Apple could make their own Ultra 3, 5, 7 and 9 line-up or are already making it.

 

 

 

 

 

 

[LAwLz]

18 minutes ago, HenrySalayne said:

- Ultra - non-trademarkable (because it's already in use by many other brands)

I don't understand your arguments.

Are you implying that a rebrand is bad if you can't trademark it? If you are, why do you think that way?

Also, are you sure that Intel can't trademark "ultra"? There were plenty of brands using that word before Apple, yet Apple managed to trademark "M1 Ultra". I don't see why Intel wouldn't be able to trademark "Intel Ultra" or whatever they end up calling their processors.

 

I also don't think the whole "AMD and Apple could make their own Ultra 3, 5, 7 and 9 line-up" is a good argument because they wouldn't want to do that anyway. Or well, I am not sure about AMD who seem to like trying to copy Intel's names (X399 chipset, R3 R5 and R7 CPUs, Phenom 940 to compete with the i7 940 released earlier and the list goes on), but I doubt other processor makers would do it.

I don't think it would be Intel that gets egg on their face if they release an "Ultra 3, Ultra 5 and Ultra 7" later this year, and then next year AMD also releases "Ultra 3, Ultra 5 and Ultra 7". But this is assuming that they will even name their products "Ultra 3, Ultra 5 and Ultra 7", which we currently have no indication of them doing and it's purely speculation.

[Senzelian]

Yes, that's what we needed; more things to be called "ultra".

[Kisai]

1 hour ago, porina said:

A problem here is that you seem to want the model number to mean everything, and it's not going to happen unless you go towards a memory chip type arrangement which is going to be even less comprehensible to the random person on the street getting a new PC.

 

No, I want the model number to be reflective of where it is horizontally and vertically in the product list.

 

If I have 10 generations of productions, the same "model suffix" should indicate where it is performance-wise. That "IS" technically true with the GPU's and CPU's, however it doesn't indicate the actual thing that happened. We didn't get "equal" power chips, we just has Intel, AMD and Nvidia slide the scale upwards (eg adding i9, 3090, 4090, etc) rather than keep the TDP equal. The "K" part in the i9, i7 and i5 all have the same 125w base TDP.

 

But go back to before the i9 was introduced. (Coffee Lake 8th gen) What's the top base TDP? 95w in the i7 and i5 K parts. These do not align horizontally. The Non-K part, however has remained at 65w? How'd they do that? 13th gen runs at 2Ghz, 8th gen runs at 3.2Ghz. Okay, but are they equal in performance?

That 5.6% difference between the non-K and K 13th gen and 3.3% difference between 8th non-K and K is significantly wider, but you'd need twice the cooling for only a 2.3% gain.

 

 

Seems like if the spec had always been about the TDP, then stores could simply point to the smaller cooler and go "that one saves energy" and most people would just buy that one instead of spending the extra money on a K or KF model that extra $15 or $25 gets them.

 

But nickel and dimes here, if you were buying just go "look how fast this goes" you'd not be buying a pre-built in the first place.

 

 

1 hour ago, porina said:

The model numbers work if you look at them for what they are: at most an indicative position within the stack of a generation. You don't need to know the exact specs but the current i9 will be generally "better" than a current i7. Yes, it gets complicated when you throw generations into the mix, but I don't think there's a good way around that.

 

But it doesn't. Like look at the GPU's where you sometimes have a lower tier part performing higher than a higher tier part simply because of the memory size. Even a lot of pre-built systems skimp and only put one memory module in, despite that leading to a substantial nerf in performance.

 

In my "perfect" version of Intel branding and part numbers. Intel would nix all the part numbers and just use the number of P cores. So Core Ultra 8 for 8 core systems and Core Ultra 6 for 6 P cores. That is better than the status quo of no correlation between the number and anything in the product. As for K, non-K or T parts, there is no reason why the K part needs to exist except to enthusiasts, so sell the K part to retail, and sell the other two to Dell and other OEM's who are not permitted to sell the K and KF parts except in a computer that has been engineered for the maximum TDP, which Dell has so far sucked at doing.

 

 

[HenrySalayne]

18 minutes ago, LAwLz said:

Also, are you sure that Intel can't trademark "ultra"? There were plenty of brands using that word before Apple, yet Apple managed to trademark "M1 Ultra".

Apple trademarked a single SKU name, Intel wants to trademark the entire line-up. They have to trademark the name combination "Intel Core Ultra" or "Intel Ultra". In contrast, "Ultrabook" is an Intel trademark that can stand on its own (without the addition of Intel).

It's generally beneficial to use a novel name and not a label everybody slaps on everything.

[LAwLz]

4 minutes ago, HenrySalayne said:

Intel wants to trademark the entire line-up.

Source?

We don't even know what the new lineup will be called. I think people need to calm down and stop making a bunch of assumptions based on a single leaked name.

 

5 minutes ago, HenrySalayne said:

In contrast, "Ultrabook" is an Intel trademark that can stand on its own (without the addition of Intel).

And why does this matter?

 

5 minutes ago, HenrySalayne said:

It's generally beneficial to use a novel name and not a label everybody slaps on everything.

Why does it matter? Seems like Apple are doing just fine, and so did Intel with for example "i5", which I don't think was a trademark either.

[darknessblade]

23 hours ago, leadeater said:

So Core i9 is going Plus Ultra? *intense BGM starts playing*

If they can't beat the Ryzen CPU's in performance/power consumption:

 

 

[CommanderAlex]

4 hours ago, LAwLz said:

I feel like you and many other people are jumping to conclusions.

We barely know anything about this rebrand, and yet people are assuming it will be a certain way and saying that's a bad way of doing things. Maybe people should hold off on judgment until we actually know what the new naming looks like?

I just cannot see why Intel would rebrand their IP from "Core i"; which makes it unique and something that anyone can recognize is Intel, to Core Ultra. Ultra is being used across the tech industry, Apple M1 Ultra, Samsung Galaxy S22 and S23 Ultra,  just like how many "X" AMD and Nvidia need to use now in their products. 

 

Apple and Intel broke up their relationship because Intel's processors, well they have been holding Apple products back and now make use their own Apple Silicon processors, where "Ultra" is one of the first marketing I've seen to come. Why the hell would Intel want to market their products "Ultra" now when they cannot even compete on a level with Apple? It also makes Intel look desperate by renaming them Core Ultra processors, where 17 years of "Core i" IP will be thrown out the door for some products that will be getting a rebrand. 

[porina]

1 hour ago, Kisai said:

No, I want the model number to be reflective of where it is horizontally and vertically in the product list.

[snip]

1 hour ago, Kisai said:

In my "perfect" version of Intel branding and part numbers. Intel would nix all the part numbers and just use the number of P cores. So Core Ultra 8 for 8 core systems and Core Ultra 6 for 6 P cores. That is better than the status quo of no correlation between the number and anything in the product. As for K, non-K or T parts, there is no reason why the K part needs to exist except to enthusiasts, so sell the K part to retail, and sell the other two to Dell and other OEM's who are not permitted to sell the K and KF parts except in a computer that has been engineered for the maximum TDP, which Dell has so far sucked at doing.

This is why I don't see a major problem with the existing arrangement. It's just a relative positioning. In your example, what do they call different generation 8 P cores? Do E cores not matter at all now? This is over thinking it. It's a name with some relative marketing positioning. It isn't a direct performance scale.

 

Intel could eliminate K by following AMD's lead and unlocking (most) parts by default. System TDP is a system builder choice, and Intel allows that to be varied from base (what Intel CPU spec is) upwards even on locked CPUs.

[HenrySalayne]

1 hour ago, LAwLz said:

Source?

Talk about laziness. You could find a source a trivial distance away from my post:

 

1 hour ago, LAwLz said:

And why does this matter?

"Get your new Incel Ultra 9000 CPU. "

[LAwLz]

5 minutes ago, HenrySalayne said:

Talk about laziness. You could find a source a trivial distance away from my post:

Sorry, but I don't watch videos from LMG because I think they are usually bad and full of factual errors.

But I did decide to look at this particular video and as I expected, there is no source for your statement. In fact, the video you linked to seems to contradict your own post.

 

You said that Intel wants to trademark the entire line-up, while the video you use as a source says that we don't know, but some websites are speculating that they won't rebrand the entire line-up to "ultra", which is what I mentioned in my earlier post as well.

https://youtube.com/clip/UgkxHK-bPCcZd1dkKkw6ANaU7MF_1XsuckRZ

 

Please dial back on the hostility.

 

 

  

15 minutes ago, HenrySalayne said:

"Get your new Incel Ultra 9000 CPU. "

I don't get it. What are you trying to say?

[LAwLz]

52 minutes ago, CommanderAlex said:

I just cannot see why Intel would rebrand their IP from "Core i"; which makes it unique and something that anyone can recognize is Intel, to Core Ultra. Ultra is being used across the tech industry, Apple M1 Ultra, Samsung Galaxy S22 and S23 Ultra,  just like how many "X" AMD and Nvidia need to use now in their products. 

I don't see how it is any less unique than "Core i". There are plenty of companies that use "i#" as their product names, and I have fond memories of people saying Intel was copying Apple when they started going with the "i" name as well. It's just that people have gotten used to it now. I don't think there is anything that is unique about Intel's current naming scheme, especially not after AMD started calling their processors "R3, R5, R7, R9" and companies like BMW calling their cars "i3".

 

By the way, it is funny to go back and look at comments from when Intel introduced the "core i" series. It's funny seeing people back then complain that the "core i" naming was so complicated to understand, and now in this thread people don't want it changed.

Spoiler

 

 

As I said earlier I am not a huge fan of "ultra", but I think we should wait until we see the entire line-up of processors and what they call them before making judgments. So far we only have the leaked name of a single SKU that may or may not be the final name, as well as confirmation from Intel that they will make changes to their naming. Everything else, including the idea that they will replace the "core i" for "core ultra" on all their processors, is just speculation.

 

Remember, what has been confirmed is that name changes are coming. What has NOT been confirmed is that Intel will start calling all their processors "ultra".

[Kisai]

1 hour ago, porina said:

[snip]

This is why I don't see a major problem with the existing arrangement. It's just a relative positioning. In your example, what do they call different generation 8 P cores? Do E cores not matter at all now? This is over thinking it. It's a name with some relative marketing positioning. It isn't a direct performance scale.

Again, the i-series arrangement just isn't meaningful. Back before Core, when we had the Pentium stuff, the number was the actual clock speed, and AMD/Cyrix basically put "PR" for "Performance Relative to a Pentium"

 

The actual clock speed has never mattered for a product except for the highest tier product, and if you could get more performance from more cores at less speed for less energy, that is preferable as long as software (read, not games) is tuned that way.

 

 

1 hour ago, porina said:

Intel could eliminate K by following AMD's lead and unlocking (most) parts by default. System TDP is a system builder choice, and Intel allows that to be varied from base (what Intel CPU spec is) upwards even on locked CPUs.

I'd rather they just TDP lock it in a way that it won't self-destruct. When it hits 95% of the TDP drop it to the base clock until the TDP drops enough that the overclock won't kill it. Put a better cooler on it, you get a higher clock speed overall.

 

But again, marketing-wise, What Intel really has right now is 5 tiers of chips that are sharing 4 labels

 

Again look at the last 3 digits:

 

900  i9

700 i7

600 i5

500 i5

400 i5 (previously i3)

100 i3

 

Is Intel so embarrassed that they have 6 product tiers that they stuff all the mid and low end parts into i5 rather than the i3?

 

Never mind the Xeon E parts which are pretty much the exact same chips with a different part number and ECC support.

[pas008]

lol this topic

 

hey min requirements are i5 cpu

 

hey i have i5 650 why doesnt it work? lol

 

geez its tech its tech it should be going forward was like 10yrs overdue, rebrand should come with new arch especially or they just use their codenames as damn brand

 

 

should always be separating old from new

 

[NumLock21]

We need more YouTube channel added to Linus's arsenal of YouTube Channels

LinusTechTips Ultra

Techlinked Ultra

Techquickie Ultra

lol

[porina]

1 hour ago, Kisai said:

I'd rather they just TDP lock it in a way that it won't self-destruct. When it hits 95% of the TDP drop it to the base clock until the TDP drops enough that the overclock won't kill it. Put a better cooler on it, you get a higher clock speed overall.

I'll leave you with your ideas but this part is just plain wrong. TDP is not the power consumption of the CPU. By definition TDP and base clock are linked. If you have a cooler rated at TDP you are guaranteed to get at least base clock. If you limit power below TDP rating, you may not get base clock depending on how stressful the load is. Better cooler = more clock is how it works right now (if you're cooling limited, other limits still apply).

[leadeater]

5 hours ago, Kisai said:

We didn't get "equal" power chips, we just has Intel, AMD and Nvidia slide the scale upwards (eg adding i9, 3090, 4090, etc) rather than keep the TDP equal. The "K" part in the i9, i7 and i5 all have the same 125w base TDP.

What the heck does TDP matter for? If the next generation product needs to have a higher TDP it's still the next generation product that replaced the old. Just because the TDP increased doesn't make it not the next xx900K etc.

 

If you were talking about the confusion between a 1360P and a 1365U then I'd get that, that is a little confusing if you don't know P and U denote entire product family/class and the similarity between the model numbering doesn't in any way mean they are comparable across P and U products. That is a problem.

 

1 hour ago, Kisai said:

Back before Core, when we had the Pentium stuff, the number was the actual clock speed, and AMD/Cyrix basically put "PR" for "Performance Relative to a Pentium"

Not really. Pentium 4 transitioned away from clock frequency as part of the model name to generic model numbering. Both naming schemes exist within the Pentium 4 product generations.

 

For example:

• Pentium 4 519J
• Pentium 4 HT 571
• Pentium 4 HT 672

etc

[Kisai]

20 minutes ago, leadeater said:

What the heck does TDP matter for? If the next generation product needs to have a higher TDP it's still the next generation product that replaced the old. Just because the TDP increased doesn't make it not the next xx900K etc.

 

Cause a i7-13700K and a i7-13700T is also a i7-13700HX and a i7-13700H in a laptop. Same with a desktop RTX 3060 being equal to a RTX 3080 in a laptop.

 

The desktop numbers reflect one thing, the mobile uses the exact same numbers but the performance is explicitly crippled because it's a mobile device.

 

The best thing Intel can do is stop conflating the mobile parts with the desktop parts, just like they don't conflate the Xeon E parts with the desktop parts despite the chip dies being the same.

 

At at least the Xeon has some logic to it

And that we consider too busy because the Performance level is meaningless. Those approximate to the same digit in the i-series 100's position. Does it matter that something is the 21st, 22nd or 23rd generation chip if the performance "8" performance is unchanged?

 

 

20 minutes ago, leadeater said:

If you were talking about the confusion between a 1360P and a 1365U then I'd get that, that is a little confusing if you don't know P and U denote entire product family/class and the similarity between the model numbering doesn't in any way mean they are comparable across P and U products. That is a problem.

 

This is just another reason why Intel's product names make no sense. They silo'd the products for servers, workstations, desktops, laptops, ultra portables, and other junk not deserving of a name.

 

I'd rather the parts either be labeled by the TDP, or labeled by the core configuration. The "all core turbo performance" is absolutely meaningless when you will never be using it that way.

 

[leadeater]

1 hour ago, Kisai said:

Cause a i7-13700K and a i7-13700T is also a i7-13700HX and a i7-13700H in a laptop. Same with a desktop RTX 3060 being equal to a RTX 3080 in a laptop.

That's still little to nothing to do with TDP itself. The is problem with the suffixing and not a more clear product type/family model name.

 

As for the GPUs the official naming from Nvidia is RTX 3080 Laptop. It has not and has never been just RTX 3080 so it's absolutely nothing to do with TDP there at all. It's only through laziness that we the consumers and tech people drop the 'Laptop' part of the name since it's redundant and implied with talking about.. a laptop. It's in the correct and official product name so there is no confusion there at all.

 

I certainly do not buy the argument that a RTX 3080 and a RTX 3080 Mobile should be the same specifications in every way, certainly not TDP.

 

1 hour ago, Kisai said:

I'd rather the parts either be labeled by the TDP, or labeled by the core configuration. The "all core turbo performance" is absolutely meaningless when you will never be using it that way.

Look TDP is entirely irrelevant, like completely. It's just a product specification and has nothing at all to do with making it a xx900 over a xx700 etc. It's just a thing and it can change generation over generation and it does. Just like number of cores change or cache or frequency or fabrication node. It's just 'a thing', it's simply not as important as you want to make it.

 

1 hour ago, Kisai said:

And that we consider too busy because the Performance level is meaningless. Those approximate to the same digit in the i-series 100's position. Does it matter that something is the 21st, 22nd or 23rd generation chip if the performance "8" performance is unchanged?

The performance is not unchanged. Performance level is a sub rating within a generation and does not span generations. You're not interpreting the model numbering correctly.

 

A E-2386G and a E-2186G do not have the same performance, single or multi core. It's not even that close.

 

It only gets worse as you go up the Xeon product family in to Gold and Platinum, which uses a different naming scheme. There are also situations like 3rd Generation Scalable which has 10nm (Ice Lake) and 14nm (Cooper Lake) in the same model numbering generation. So entirely different CPU core architectures.

 

1 hour ago, Kisai said:

The best thing Intel can do is stop conflating the mobile parts with the desktop parts, just like they don't conflate the Xeon E parts with the desktop parts despite the chip dies being the same.

How is E-2186G (desktop), E-2186M (Mobile) any different from i7-13700 (desktop), i7-13700H (Mobile)?

[WallacEngineering]

On 5/1/2023 at 4:43 AM, leadeater said:

So Core i9 is going Plus Ultra? *intense BGM starts playing*

Dang, I was gonna do that joke

 

I think the new branding would be better if the did it like this.

 

i3 - i3

i5 - i5

i7 - Rage i7

i9 - Ultra i9

 

Or something like that. I don't see out the "Ultra" mentality fits with lower end CPUs.

[vertigo220]

On 5/1/2023 at 9:48 AM, LAwLz said:

I saw someone in the comments mention that it might be like this:

Intel Core i3 = Intel Core Max

Intel Core i5 = Intel Core Ultra

Intel Core i7 = Intel Core Extreme

Intel Core i9 = Intel Core Extreme+

This is a perfect example of why such word-names can be problematic. To me, max should be above ultra and extreme, since it's, well, max. Ultra and extreme can be debated as to which should be above the other. So hopefully they don't choose to go this route. The bronze/silver/gold/platinum is better, but still not great. It's longer, it doesn't allow as many gradations, and it's already used to an extent by their Pentium line, and therefore may cause confusion between the two. IMO, numbers are the best in every respect, and now that they're in double-digits for the leading number(s), by the time the size of the SKU would increase again, we probably won't even be using x86 CPUs anymore. If they wanted to differentiate and possibly reset, they could (and maybe will) do something like i3+ or x3 or z3 or whatever.

 

On 5/1/2023 at 10:16 AM, MageTank said:

In all seriousness, the biggest problem is the competing companies trying to win the "bigger number wars". Intel had the Zx70 chipsets, AMD releases the Xx70 chipsets. Intel makes Zx90 to counter in consumer. Same thing happened when Intel's HEDT chipset was X299, AMD release X399, lol. We see it on the graphics segment between AMD and Nvidia, and processor naming conventions was very similar during the Phenom II era when Nehalem launched. Intel launched Nehalem Core i7 940, AMD launches the Phenom II 940 Black Edition just 2 months later.

 

Bonus points if AMD and Intel agree to stop using similar chipset names/numbers to one-up each other. Imagine trying to buy a modern board with AMD and Intel both using LGA now and you personally don't know the difference based on names alone. Must be a nightmare for the newer folks getting into PC building;.

I personally think if they're same generation and competing, especially if they're close to each other, it makes sense to have the numbers be the same/similar, as it helps less knowledgeable consumers recognize that they're competing, similar options. And I'm sure that was the point of them doing that. And especially considering Intel's history of anti-competitive behavior toward AMD, AMD had to take steps it may not have otherwise taken in order to compete.

 

21 hours ago, NumLock21 said:

Ultra should only be at the very top, naming everything with Ultra just defeats the meaning behind it. A Intel Celeron Ultra is nothing to be bragging about.

As I mentioned above, IMO ultra would be at least below max. But while we don't know yet what their exact plans are, it could very well be they are using it for the exact reason you mention, which is that a Celeron Ultra is still garbage, but people may see the "Ultra" and think it's really good.

 

19 hours ago, Arika S said:

OK, but why?

 

So Intel is never allowed to rebrand their stuff ever?

 

Was amd more or less pathetic when rebranding from athlon fx to fx and then fx to ryzen?

Fair enough, but with AMD's rebranding, as well as Intel's previous rebranding, it's generally, if not always, been alongside major architecture changes, i.e. going from a 2- or 4-cylinder sedan to a 4/6/8-cylinder sports car, not from a 4/6/8-cylinder sports car with 350HP to one with 380HP (and worse gas mileage). IMO, Intel needs to do something much more innovative than they have been to justify a rebrand. In fact, if anything, it should have been done when they switched to P/E cores.

 

8 hours ago, LAwLz said:

I also don't think the whole "AMD and Apple could make their own Ultra 3, 5, 7 and 9 line-up" is a good argument because they wouldn't want to do that anyway. Or well, I am not sure about AMD who seem to like trying to copy Intel's names (X399 chipset, R3 R5 and R7 CPUs, Phenom 940 to compete with the i7 940 released earlier and the list goes on), but I doubt other processor makers would do it.

I don't think it would be Intel that gets egg on their face if they release an "Ultra 3, Ultra 5 and Ultra 7" later this year, and then next year AMD also releases "Ultra 3, Ultra 5 and Ultra 7". But this is assuming that they will even name their products "Ultra 3, Ultra 5 and Ultra 7", which we currently have no indication of them doing and it's purely speculation.

See my response to MageTank re AMD "copying" Intel's naming.

[MageTank]

19 minutes ago, vertigo220 said:

I personally think if they're same generation and competing, especially if they're close to each other, it makes sense to have the numbers be the same/similar, as it helps less knowledgeable consumers recognize that they're competing, similar options. And I'm sure that was the point of them doing that. And especially considering Intel's history of anti-competitive behavior toward AMD, AMD had to take steps it may not have otherwise taken in order to compete.

I disagree. Having the naming be so similar is only going to further confuse amateur system builders into buying an AMD motherboard with an Intel processor, or vice versa. AMD intentionally named their chipsets to be nearly identical to Intel but putting the numbers higher because most consumers have the pre-conceived notion that higher = better.

 

Also, what anti-competitive behavior are you referring to in this instance?

[vertigo220]

6 minutes ago, MageTank said:

I disagree. Having the naming be so similar is only going to further confuse amateur system builders into buying an AMD motherboard with an Intel processor, or vice versa. AMD intentionally named their chipsets to be nearly identical to Intel but putting the numbers higher because most consumers have the pre-conceived notion that higher = better.

 

Also, what anti-competitive behavior are you referring to in this instance?

I guess I can see that, though anybody building their own computer should at least know enough to not mix AMD and Intel CPUs and motherboards, and if they don't, sorry, but that's on them, and they should have done just a tad more research.

 

It's well-established and can likely be found easily with a bit of searching, so I don't want to get into it here. But there's a reason why laptops have historically been almost entirely Intel, and even now, with AMD being regarded by most as superior to Intel for a few years now, the majority of laptops continue to use Intel vs AMD.

[cmndr]

On 5/1/2023 at 8:48 AM, LAwLz said:

I don't really get how someone can be confused by this:

7700

8700

9700

10700

11700

12700

13700

 

I know someone who had an i7 930 who was fine with its performance but the system died. 
He was on a budget and was kind of shocked I'd suggest an i3 12100 since an i3 was "slower" than an i7. 

The i3 12100 has ~2x the IPC, runs at ~30% higher clock and supports a bunch of hardware acceleration the ancient chip from 2010 does not. It's generally 2-3x faster.