Intel Explains Brand Strategy with Core

[lemonadesoda]

Do you happen to think that Lynnfield doesn't use QPI as a system interconnect?

Let me split a hair with you: Lynnfield uses QPI for intraconnect and DMI for interconnect.

Yeah, but this naming scheme opens up the possibility of calling a car built on the 3 series platform, a 7 series. And I still say this is a marketing blunder. No cpu on the 1156 socket should be i7. I think i7 should be for triple-channel and QPI only.

Yes, it would be like calling the BMW M3 a 7 series.

Intel have screwed this up a bit in "the enthusiast" community. However, most consumers havent even heard of the terms i7 or Nehalem yet, so perhaps Intel can "manage" the situation. It's no excuse though... for a badly thought out marketing plan. Fire the lot of them.

 

[Hayder_Master]

Arrrrr , i3 thanx for nothing intel

 

[btarunr]

Let me split a hair with you: Lynnfield uses QPI for intraconnect and DMI for interconnect.

It does not matter. All QPI does on Bloomfield is connecting the processor to the X58 northbridge. All it does on Lynnfield is connecting the CPU die to the northbridge die. X58 continues to use DMI to connect with the ICH10R. Architectural difference? Null.

Does it matter how big the green rectangle is? It can even gulp down the ICH/PCH and become an SoC. It still wouldn't make for an architectural difference. The QPI link on Lynnfield is user controllable, just like the one on Bloomfield. Therefore, there is no difference in the roles they are playing, and it shouldn't matter where it is present. Lynnfield does use QPI, and in the same way Bloomfield does. In no way is Bloomfield superior to Lynnfield when it comes to implementing QPI.

 

[Sihastru]

Actually because the NB die + CPU die are on the same package for Lynnfield this could translate into a small advantage when comparing latencies. So in single video card configurations (1 x PCIe 16x), Lynnfield might give you a just few extra frames per second.

 

[PCpraiser100]

Actually because the NB die + CPU die are on the same package for Lynnfield this could translate into a small advantage when comparing latencies.

That will be so until they are overclocked. This is why AMD hasn't done this, as this kind of design, or advantage, should be only taken to platforms that require minimal setup that seem careless about heat, like cell phones and game consoles. This is not recommended for media towers however should suffice with good old mATX and ATX towers. Taking this to the notebooks will remain decent until you either burn through your favorite pants with the heat or run out of battery before noticing.

 

[jagl4d]

SKU Nightmare

One hell of a nightmare for systemboard makers and retailers. Even recycling So much all for going green.
People looking for midrange to high performance boards and prociess will have to shell out more $$$. Nice intel, just nice!

 

[locoty]

and Intel CPU now like GPU in the terms of naming scheme

There is i3, i 5, i7

i'am not surprised that there will be i9, i11, i13 and so on

 

[Sihastru]

There will be an i9. I am hoping for an i1, since these processors are too powerful.

 

[mdm-adph]

Wow -- reading all this makes me glad I use AMD chips.

I don't care if they're slower -- what Intel has done looks like a massive, expensive fucking headache.

 

[jessicafae]

I hope Intel did extensive marketing research on this. there is a small possibility this might not work exactly like they hope.

Given this chart I bet this is just marketing so they can eventually say "i5 is approximately equal to PhenomII and i7 is better" or something similar to try to "dumb down the consumer". "why is this laptop $300 more? it has an i7 which is faster..." . Why did they place the Clarksfields mobile processor in i7 and there are no mobile processors in i5? I bet it will be slower than the current i7 920.. it is a mobile processor after all with thermal limits.

Of course anyone can play the FUD/simplification game... if this accelerates it could lead to a bit of backlash. People don't need to upgrade, computers are generally fast enough, people are looking at computers more and more like a cellphone, music player, or TV (something that just works and does xxx). The netbook / iphone / ps3-xbox360 factor.

Consumer: Why do I need a faster computer?
Sales person: Well you need more power to video encode your Bluray movies down to 480x320 resolution for your ipod
Consumer: hmmm

 

[Wile E]

Wow -- reading all this makes me glad I use AMD chips.

I don't care if they're slower -- what Intel has done looks like a massive, expensive fucking headache.

How is AMD's naming sceme any less confusing to the consumer?

I'll tell you right now, it isn't.

 

[mdm-adph]

How is AMD's naming sceme any less confusing to the consumer?

I'll tell you right now, it isn't.

Not talking about the naming scheme -- I'm talking the possibility of all these sockets.

Right now, I can run anything from AM2, to AM2+, to AM3 on my board. To me, that's much, much easier. That's why I said it makes me glad that I use AMD chips.

But please -- return and tell me how I'm wrong, how they're the same, blah blah blah.

 

[TheMailMan78]

Wile E take it from me. Arguing with mdm-adph is like arguing with water. He just flows around your rocks of logic and common sense.

 

[Wile E]

Not talking about the naming scheme -- I'm talking the possibility of all these sockets.

Right now, I can run anything from AM2, to AM2+, to AM3 on my board. To me, that's much, much easier. That's why I said it makes me glad that I use AMD chips.

But please -- return and tell me how I'm wrong, how they're the same, blah blah blah.

Can't run any Athlon 64/64 X2, Phenom I's or PII 940's in an AM3 board, for example, as they lack DDR3 controllers. Sometime in the relatively near future, you won't be able to run new AMD cpus in AM2+ or earlier boards, as they will lack a DDR2 controller as well. And this isn't the first time with multiple socket for AMD either. Remember the 939 days? 754, 939, and 940 all at the same time.

AMD and Intel both are guilty of similar things throughout the history of both companies.

 

[[I.R.A]_FBi]

If you dont like intel dont be all up in their grill is all

 

[mdm-adph]

Wile E take it from me. Arguing with mdm-adph is like arguing with water. He just flows around your rocks of logic and common sense.

Your rocky logic and common sense?

What -- you're running an AMD chip. Stop sticking up for them. You're either with us or against us.

Can't run any Athlon 64/64 X2, Phenom I's or PII 940's in an AM3 board, for example, as they lack DDR3 controllers. Sometime in the relatively near future, you won't be able to run new AMD cpus in AM2+ or earlier boards, as they will lack a DDR2 controller as well. And this isn't the first time with multiple socket for AMD either. Remember the 939 days? 754, 939, and 940 all at the same time.

AMD and Intel both are guilty of similar things throughout the history of both companies.

Yes, the 754/939/940 days sucked. However, like I said, for me, right now, AMD seems like the much, much easier/cheaper choice. There's no need to be offended because you run Intel chips.

 

[Wile E]

Your rocky logic and common sense?

What -- you're running an AMD chip. Stop sticking up for them. You're either with us or against us.



Yes, the 754/939/940 days sucked. However, like I said, for me, right now, AMD seems like the much, much easier/cheaper choice. There's no need to be offended because you run Intel chips.

I'm not offended. And there's no need to defend AMD in an Intel topic, just because you run an AMD chip.

 

[jagl4d]

Intel could have just made their i5s & i3s similar to the i7 layout with No Connects & their ICH chips reconfiguring accordingly. That strategy could well sink AMD deeper than ever.