News Intel announces Atom x3, x5, x7 branding

Why don't they just tell me how many cores at what speed... you know, like the old days....

 

A good idea in principle but I can't help but think that they should have used a different letter than 'x' which somehow leads me to think in mathematical terms and that the Atom x3 will then be a 3 core chip where as the x7 is a 7 core chip for example. I know that this is not the case but my brain can't help but get diverted for a second or two. I would have thought using 'i' would be fine as you still have Atom, Celeron, Pentium and Core to differentiate between levels of chip.

 

Because these days there is far more to a chip than just core count and clock speed? For example with the Core i5 and i7 chips, if you look at the two high end 'K' versions the only real difference (apart from 2MB on L3 Cache) is that the i7 features HT where the i5 does not. Technically core count is the same and clock speed is often very similar between i5 and i7 so without knowing more specific details it is hard to tell why the i7 is worth a significant amount of extra money just by the details you are requesting.

 

@SchizoFrog
I agree about the whole "x" thing. But intel is notorious for AWFUL product names. Pentium, Celeron, Xeon, and Itanium were excellent names, but have been living for over 20 years because they just can't let go or think of anything better. "Atom" and "Quark" are nice product names, but at least Atom hasn't been all that impressive.

Anyway ranting aside, I think its about time intel decided to differentiate between their low-power range. Considering they're trying to directly compete with ARM, it makes sense that they need a more specific naming scheme. I just really hope Intel finally ditches the Pentium name. I know they won't, but what is now Pentium (or maybe Celeron) could always be Atom x7.

 

One small error: "conpare"

I couldn't help but notice it.

 

Also greed on x[Whatever]... 'A' and 'E' would cause confusion with AMDs APU wares [or intel's Xeon line], so i'd probably have gone for n[whatever]... But even without the 'Atom', 'Core' or w/e it should be clear where in the pecking order it sits...

 

There is.. Give me the codename instead then. Nice and easy to compare then.. I mean, I know for a fact that 4 Haswell cores should be faster than an equally clocked Core 2 Quad for example

 

I don't understand your issue as you're explaining it. Forget Celeron and Atom as we all agree that there has been a problem there which is also why Intel are trying to do something about it. The way I understand Intel's strategy is like this: (although I admit there are a few cross over details but as a rule it works)
Pentium - Dual Core, i3 - Dual Core with HT, i5 - Quad Core, i7 - Quad Core with HT. As for understanding which generation is better, that is what the serial numbers are for: 3xxx - Ivy Bridge, 4xxx - Haswelll... with K being the unlocked multiplier chips and the xx90's being the updated revisions.

With such an extensive and broad spectrum of products and several generations of those items on the retail market at any one time I don't any other way than to have a rather broad product coding system. Unless someone can point out my errors, it all seems rather straight forward to get the hang of it. Currently, it certainly seems to make more sense than AMD who seem to lose the plot to such an extent I actually stopped following their product releases...

 

lol I wish it were that simple but unfortunately that's not true. The different product names are largely based on processor instruction sets and abilities. That's why it's [usually] stupid to get a laptop with an i7 - there are i3s, i5s, and i7s for laptops that are dual cores with HT. The only major differences between them are the GPUs, clock rates, sometimes cache, and instruction sets you're probably never going to use. Sometimes the NB is tweaked a little but on a laptop that hardly ever matters.

To the average user, intel's naming scheme is utterly useless. To most enthusiasts, it's still useless, as even you (who I trust is adequately well-informed) didn't get intel's naming scheme right. Note that I am also greatly generalizing. Since intel's products change so much over the years or over different platforms, it's kind of hard to explicitly state what is different about each of their products.

AMD's naming scheme is simple, the letter resembles the "family" of product (so for example E series is power-efficient APUs, A series is the standard desktop APU, FX series is CPU, R series is GPU) and the number attached to that letter resembles the "class". Any number after that gives you the specific name. BTW, I'm putting this in my own words, in case you're wondering "why have I never heard this before?"

EDIT:
I don't actually like AMD's new naming scheme, I just think it's easier than Intel's. The only thing truly idiotic about it is how they have those R7 SSDs. That's got to be insanely confusing to some people.

 

OK, I'd like to separate between laptop and desktop parts especially as I did mention 'retail' parts and laptop CPUs are not available to the public. Although I do agree that laptop parts can be misleading and that they shouldn't be directly compared to desktop parts (it's much the same in the GPU world as a 970M is nothing like a desktop 970) I'd question your comment that laptop i7's are or can be dual core albeit with HT (if you can show me otherwise, fair enough). Every one I have looked at over recent months or even the last couple of years matches that of the desktop parts as in quad core with HT. As exampled below:

Laptop
http://www.scan.co.uk/products/156-...50gb-hdd-2gb-nvidia-geforce-gtx-860m-dvdrw-wi

CPU Details
http://ark.intel.com/products/78930/Intel-Core-i7-4710HQ-Processor-6M-Cache-up-to-3_50-GHz

So once we look at desktop parts only, I am not sure whether I still 'got it wrong' or not?

EDIT: OK, while I haven't seen any modern laptops with them in, I have seen dual core i7's referred to so fair enough. However, Intel does differentiate between quad and dual core with different letter sequences so I am back to not seeing much of an issue. Personally I think that some confusing is good, it SHOULD lead people to ask questions and LEARN what they are buying rather than buying something blindly.

 

I both agree and disagree. It is good to encourage people to learn before they buy, but this is hardly a good way to make people learn. It also makes the naming scheme completely obsolete - the purpose of these names is to gauge roughly where the performance should be, while the number of the product tells you exactly where it stands. But in some cases you might be able find an i3 outperforming an i5, or a low-end i7 that is less than half the performance of a more expensive i7. Setting core count, HT, and GPU size aside, the architectural differences between an i3, i5, and i7 are so minimal that the names are overall useless. They're supposed to reflect the overall performance of the chip but on what metric? When it comes to intel, I completely ignore the series names because they're just misleading, and frankly, a marketing joke.

 

But X sounds more awesomer than I...

 

@Schmidtbag
Well you say they are misleading and useless. I disagree. You want to dissect and dismiss relevant and integral features that don't suit your arguments. You seem to believe that there is no performance differentiation between i3, i5 and i7's and again I disagree. With regards to Pentiums and upwards I believe I understand where they all fit in across the board and also across multiple generations although you have already told me that I misunderstood the naming system and was incorrect. However, you have failed to educate me with correct examples, nor have you suggested anything better and simpler. As such, even if you are ultimately correct I am never to know better and so I shall muddle on in my confused manner.

As it stands I don't see anything inherently wrong with Intel's naming system. Could it be made more simple? I am sure it can, but then they could also just make five CPUs to cover the full range but they won't and so with many more products comes the need for a more complex naming system and there is far more to a CPU than just a benchmark score.

 

Admittedly never heared of "Quark", but I always liked the impressively low use of power by atom processors.

 

@Gareth
To be fair, Quark isn't really an appealing product to most people. It isn't the most practical product out there (at least when you include price point and the setup process). The only major benefits of Quark over an ARM/atmel equivalent is PCIe 2.0 support. Being able to run closed-source x86 software is definitely a nice bonus but I can't think of any real-world x86 application that would be useful enough and perform well enough to be practical.

There are of course reasons to specifically want the Quark platform, but I think most people would be better off with ARM.

 

I can't explain it either, I read your news, really!!! , but no, never heared of quark.

Probably because the "Internet of Things" isn't really on my radar yet, or at least I haven't found an affordable use for it.