An early look at Phenom dual / tri core performance

[caamsa]

Yes. You can quote me or anyone on that.

Intel won't try to put 4 duals together. Intel will move to a native quad and then use two quads to make an Octo.

Maybe AMD should have stuck two dual core K8's on a die before trying to go for a native quad core.

I will be very interested in seeing how the intell native quad does. Maybe putting too many cores on one chip is a bad idea and stacking up dual cores may actually be the way to go.

 

[epsilon84]

Now, I know this has no fact behind it ... but to me it seems to be quite fair reasoning. It also makes sense that if it's fixed, and any minor errors produced are cleaned up, the performance of K10 vs. K8 could rise a couple to a few percent. Not enough a gain to challenge Intel, but then phenom could do with every ounce of performance it can get, right?

I wouldn't rule out such a possibility, but like you said, it'll only be a few % points difference, and is not likely to change the competitive landscape. What AMD really needs is to scale Phenom to 3GHz and higher. A few % higher IPC will not do much when it is 30% behind in clockspeed.

 

[TechnologyCoordinator]

Maybe AMD should have stuck two dual core K8's on a die before trying to go for a native quad core.

I will be very interested in seeing how the intell native quad does. Maybe putting too many cores on one chip is a bad idea and stacking up dual cores may actually be the way to go.

You are correct.

Native Quad @ 65nm = Fail for AMD or Intel

That is why Intel did the double cheeseburger, but it worked really well. AMD should have done something similar. Imagine if AMD could find a better way and manufacture single cores and bin them into an ultra-fast quad. I'm just throwing ideas out there. But native quad @ 65 would be a disaster for AMD or Intel.

 

[Xpyrofuryx]

AMD... I love you...

 

[epsilon84]

You are correct.

Native Quad @ 65nm = Fail for AMD or Intel

That is why Intel did the double cheeseburger, but it worked really well. AMD should have done something similar. Imagine if AMD could find a better way and manufacture single cores and bin them into an ultra-fast quad. I'm just throwing ideas out there. But native quad @ 65 would be a disaster for AMD or Intel.

Sharikou lingo is infesting your mind TC...

 

[boner]

aint that a b!tch......... and also man sean taylor died and that's another sad thing............... dont ask why i brought that here

 

[coret]

I wouldn't rule out such a possibility, but like you said, it'll only be a few % points difference, and is not likely to change the competitive landscape. What AMD really needs is to scale Phenom to 3GHz and higher. A few % higher IPC will not do much when it is 30% behind in clockspeed.

I have to agree - AMD's clock speeds are not impressing, and the power consumption & heat production going on at these clocks is painful to see. I 'upgraded' (change from core2 with 7950gt to 6000+ with 8800gts ... did it for the graphics and managed not to lose any money overall) a few months back to my current system, and I'm starting to wish I'd saved up a little more money and gone for the G0 q6600 option instead.

 

[elbert]

Are you still sticking by your 'Octo core by summer 08' prediction then? :lol:

There aren't that many unknowns actually. We already know the performance of Phenom. A shrink to 45nm will not bring miraculous gains in performance. I don't see it matching Yorkfield, let alone Nehalem.

Of course, you may have different views on that, and only time will tell who is right. 😉

We already know that Yorkfields performance isnt much better than Kentsfield. Na I just don't see Nehalem bring much higher per core performance over Yorkfield. Mostly just a twice the number of cores. The 45nm shrink for phenom will bring in much higher clocks as IBM's 45nm process will use hafnium just as Intels. We could see 4GHz from Phenom which should make it close.

http://www.engadget.com/2007/01/27/intel-ibm-announce-insulation-overhaul-for-faster-cooler-chi/

 

[caamsa]

Intel Pumpers! The end is here now!

Phenom is the 100% super-chip destroyer of evil Nazi Intellers! At its super-high clocking of GREATER than the 2Ghz it is so fast that evil Intel "Benchmarks" Cannot be used in the measuring of its fastitude!

Fact 1: Phenom is already the most popular CPU in all of the desktop history!! AMD has the sales of at least 10 BILLION in this month! This is obvious since Phenom comes in many different cores while Stupid Evil Intel only has the 2 cores or the "glue" in the 4 cores!!
3 Cores is better than the 4 cores unless AMD says that the 4 cores is better!

FAct 3: All Intel CPUs are the VAPORWARE. Oh, Intel says it makes some lie called "core 2" but here is truth: Those chips are only old AMD chips that the INTel STOLE from the AMD and then lied and said it made them. INTEL HAS NEVER MADE A SINGLE CHIP EVER!! IT ONLY STEALS FROM THE AMD!!!
AMD will always be the 10x generation leader because Intel can only steal the old AMD chips!

Penryn is the most delayed CPU in all of the history of the computing!! This is because the dumb Intellers could not steal the 10 year old AMD chips!!
Nehalem is a lie!!! Just compare: Phenom is the most widely used and tested chip in the history of all the world! Randy Allen has powerpoints proving Phenom is the bestest of AMD!! But Nehalem is just a stupid demo! INTEL VAPORWARE!!

Fact B: Intel is ALREADY BK!! Do not believe some lying pro-Intel pumper who says Intel will survive until 2009. Do not believe some fanboy who is saying that the 2Q 2008 BK will come true!! These are all Intel lies! It uses its illegal money to buy off the pumpers, but HECTOR knows that it is already the BK!!!

Fact VIII: All Intellers are the evil Nazis as the Sharikous proved! Intel people believe in evil ideas like "thinking" and "facts". All good PURE AMD peoples know that this is evil and dumb. Only the Hector is needed for thinking! Only things that the AMD approves are allowable!
Do not believe evil Intellers who talk about "choice" or "competition" These words are the stupid and bad. Since Intel is EVIL any choice that is not AMD IS THE EVIL! Only AMD must be the processor! Only AMD may be legal in the usage!
Choice is the Evil! AMD MUST RULE AND ALL OTHERS MUST DIE! Heil Hector!

(The Burninator 10/7/07)

Tell us how you really feel?

I think a lot of people feel about Intell the same way a lot of people feel about microsoft......they hate them...just like people hate walllmart.....etc.....I see their point (big businesses can smother smaller ones and use their muscle to keep others out of the business) But IMO if the product is good and the price is right you make the logical choice.

When AMD was kickin it with K8 over netburst AMD should have been more aggressive. They needed some PR to get the word out. Unless things change they will never be able to compete on an equal level with intell. Just like the mom and pop stores will never be able to compete against wallmart.

 

[epsilon84]

We already know that Yorkfields performance isnt much better than Kentsfield. Na I just don't see Nehalem bring much higher per core performance over Yorkfield. Mostly just a twice the number of cores. The 45nm shrink for phenom will bring in much higher clocks as IBM's 45nm process will use hafnium just as Intels. We could see 4GHz from Phenom which should make it close.

http://www.engadget.com/2007/01/27/intel-ibm-announce-insulation-overhaul-for-faster-cooler-chi/

Intel claims Core 2 -> Nehalem will bring an bigger jump in performance than Netburst -> Core 2, which is ~40%. Whilst I find that a little hard to believe, the IMC alone should mean better scaling in multithreaded performance, which is the last main advantage AMD has over Intel at this point.

Again, at this point this is just speculation.

 

[elbert]

Actually, Nehalem will be native for both 4 and 8 core configs.

And oh, I missed the comment about quoting. Sure, you can quote me if you want... 😉

My question to Elbert is - can we quote YOU in saying AMD will have an 8 core CPU by mid 2008? :lol:

AMD may have 8 cores by January just 2 CPU's on a 1207+ FX. You can quote me on AMD having an 8 core CPU by mid 2008 but thats not to say we can buy one. It would be released first to the 1207+ server market.

 

[epsilon84]

AMD may have 8 cores by January just 2 CPU's on a 1207+ FX. You can quote me on AMD having an 8 core CPU by mid 2008 but thats not to say we can buy one. It would be released first to the 1207+ server market.

Good then. We shall wait and see.

 

[elbert]

Intel claims Core 2 -> Nehalem will bring an bigger jump in performance than Netburst -> Core 2, which is ~40%. Whilst I find that a little hard to believe, the IMC alone should mean better scaling in multithreaded performance, which is the last main advantage AMD has over Intel at this point.

Again, at this point this is just speculation.

Are you saying Intel is claiming Nehalem will beat a C2D by 40%? If so I could see that as even the C2Q can do that. Are you saying Intel is claiming the 8 core Nehalem will beat the C2Q by 40%? Again I could see that. I am left thinking is Intel just trying to confuse us? The last advantage AMD will have is HT3 and could bring up a possible bottleneck for Nehalem. I just wonder if Intel is going to address the 4X4 threat which may surface around January and again on Nehalems launch.

 

[elbert]

Good then. We shall wait and see.

Man I hate waiting. Don't you?

 

[coret]

Uhm, just want to suggest something with Intel's performance claims for nehalem. Oh, and I'm not trying to be an AMD fanboy, just pointing out a possibility.

Do they actually say that they're talking about per-clock performance, or are they generalising?

I only ask, because most of the produced netburst were single core, with most core2 models being dual core. Single -> Dual = 100% increase.

Most Core2 models are still dual core, with the lower tier (and thus most produced to start off with) of nehalem will be Quad core. Dual -> Quad = 100%.

But unlike with Core2, they're actually planning to sandwich 2 nehalems together quite soon after launch (if not with launch) to make octo-cores, which would obviously have more than simply double the number of cores ... and thus more than double the performance increase (application depending).

I'm just pointing out, that if they haven't said that they're talking about per-clock performance (which i'm sure will be greater than core2's anyway) this performance talk could just be some clever marketing on intel's part.

 

[coret]

Are you saying Intel is claiming Nehalem will beat a C2D by 40%? If so I could see that as even the C2Q can do that. Are you saying Intel is claiming the 8 core Nehalem will beat the C2Q by 40%? Again I could see that. I am left thinking is Intel just trying to confuse us? The last advantage AMD will have is HT3 and could bring up a possible bottleneck for Nehalem. I just wonder if Intel is going to address the 4X4 threat which may surface around January and again on Nehalems launch.

You beat me to it 🙁

 

[elbert]

You beat me to it 🙁

Sorry!!! You did say it better.

 

[epsilon84]

Are you saying Intel is claiming Nehalem will beat a C2D by 40%? If so I could see that as even the C2Q can do that. Are you saying Intel is claiming the 8 core Nehalem will beat the C2Q by 40%? Again I could see that. I am left thinking is Intel just trying to confuse us? The last advantage AMD will have is HT3 and could bring up a possible bottleneck for Nehalem. I just wonder if Intel is going to address the 4X4 threat which may surface around January and again on Nehalems launch.

More specifically, Intel claims Nehalem will bring a higher IPC boost over Penryn, than Core 2 over Netburst. Since they are talking IPC, it wouldn't be due to the core count.

http://www.theinquirer.net/gb/inquirer/news/2007/10/16/idf-taipei-nehalem-real-big

So, how much will the Nehalem be faster than the already oh-so-fast Penryn, focusing specifically on the core IPC single thread performance? I asked Kirk Skaugen, Intel's Digital Enterprise Group VP and GM of Server Platforms Group.

...he didn't want to comment on the CPU core-specific performance expectations beyond the well known integrated memory controllers and interconnects. But he did say that the CPU core performance jump from the same process Core 2 (Penryn) to Nehalem would be higher than the jump for Netburst to Core 2 itself.

Seems rather hard to believe, doesn't it? If true, Nehalem would be an absolute monster, especially at 8 cores with HT.

Of course, time will tell if Intel is indeed telling the truth. A lot was made of AMD's '40% faster' claims and we all know how that turned out.

 

[coret]

Ok, with that highlighted like that, it's obvious they're not generalising like i suggested. However, it's per-core, and not per clock ...

We have to consider here that the ebil Intel may have sold the souls of their employees to the devil in exchange for uber-high clocking tech for nehalem ^^

 

[epsilon84]

Ok, with that highlighted like that, it's obvious they're not generalising like i suggested. However, it's per-core, and not per clock ...

We have to consider here that the ebil Intel may have sold the souls of their employees to the devil in exchange for uber-high clocking tech for nehalem ^^

Fair about about per core vs per clock.

Either way, be it per core or per clock, Intel claims it will be MUCH faster than Penryn, which, like I said earlier, is rather hard to believe, but time will tell.

 

[coret]

Agreed ... i'm just wandering though if they got some pentium 3's, duct-taped them together and downsized them to 45nm, and then used the devils magiks to make them run (and work) at 10GHz ^^

<sigh>

Pentium III's were cool....

 

[elbert]

More specifically, Intel claims Nehalem will bring a higher IPC boost over Penryn, than Core 2 over Netburst. Since they are talking IPC, it wouldn't be due to the core count.

http://www.theinquirer.net/gb/inquirer/news/2007/10/16/idf-taipei-nehalem-real-big

So, how much will the Nehalem be faster than the already oh-so-fast Penryn, focusing specifically on the core IPC single thread performance? I asked Kirk Skaugen, Intel's Digital Enterprise Group VP and GM of Server Platforms Group.

...he didn't want to comment on the CPU core-specific performance expectations beyond the well known integrated memory controllers and interconnects. But he did say that the CPU core performance jump from the same process Core 2 (Penryn) to Nehalem would be higher than the jump for Netburst to Core 2 itself.

Seems rather hard to believe, doesn't it? If true, Nehalem would be an absolute monster, especially at 8 cores with HT.

Of course, time will tell if Intel is indeed telling the truth. A lot was made of AMD's '40% faster' claims and we all know how that turned out.

The inquirer? Better have some salt on hand. In the old days they laughed at anyone quoting that which should remain unnamed. If true it would raise some serious problems for AMD. More the reason AMD needs to take most of this year as the top CPU's on a 1207+(FX).

 

[elbert]

Agreed ... i'm just wandering though if they got some pentium 3's, duct-taped them together and downsized them to 45nm, and then used the devils magiks to make them run (and work) at 10GHz ^^

<sigh>

Pentium III's were cool....

I think its more dire. Maybe the dreaded 486 shrink. Intel has forgot more good technology than we will ever know.

 

[epsilon84]

TheINQ is not very reliable in terms of releasing rumors and such, but this was a direct interview with a senior Intel exec, so if it turns out wrong, TheINQ isn't at fault, but rather Intel (or more specifically, Kirk Skaugen) for spreading false hype.

I've read posts from people who are Intel employees on various tech forums, and whilst they are bound by NDA, they share the same view in that Nehalem will be a massive step forward in performance.

Whether it will actually be >40% faster than Penryn 'per core' remains to be seen, but heck, if it's even 20% faster than Penryn that's already very impressive.

 

[coret]

I think its more dire. Maybe the dreaded 486 shrink.

I'd buy one just for the novelty of having a 45nm 486