THG (P)reviews "Core 2 Quadro" - aka Kentsfield!

[1Tanker]

AdvancedMicro what?

Follow-up here's the best part, this data was generated using an Asus P5W DH Deluxe Mb Rev 1.02G --- which I have Sorry for going OT.....but

@Jack. Did you get your X6800 yet, and if so, are you still going to do the FSB/BW tests using the unlocked multilplier?

 

[turpit]

DELL just bought 2 million + Athlon
COVERTING 1/3 RD of its line to AMD

Intel is cutting jobs like by the 10's of thousands....

The sky is falling, the sky is falling......
Who cares? WTF does that have to do with this thread? NADA

tell me....what kind of availability do you expect KENSFIELD to show...

Now this question seems a little more legitimate.
In response: Judging by the Core 2 Duo vs X2 5000 availability, there should be no problems with the Intel quad cores availability once they hit the market

 

[symbi0nt]

Bah come on, it's monday already. This means they should post the article!

 

[MU_Engineer]

C'mon, we all know that Intel uses a different method (UL consumer goods spec) than AMD (maximum possible disspation: Vcore * Imax) so the TDPs cannot really be compared. The maximum power dissipation of a 65W Conroe is 87 W (Vcore * Imax) while the 65W X2 EE is actually 65W. It is also funny to note that Intel rates its Xeons, or at least did with the P4-based Xeons, by 90% of Vcore*Imax instead of the 75% as per the UL spec. So a 4x4 with AMD's EE processors can draw as much as 130W while the Kentsfield can draw 174W. No wonder during Tom's testing it was in the upper 60s like the hottest of the Smithfields.

It looks like it will be a long road to having more than 4 cores. Intel's problem will be keeping power dissipation low enough and AMD's problem will be to fit it all on one die. Maybe when Intel gets to 45nm we'll start to see rumblings of 8-core chips and when Intel hits 32nm and AMD hits 45nm we'll get them.

 

[Pippero]

For the SSE, your explanation is much more sensible, I really don't know - I mean it would seem logical that if you wanted to push SSE the first step would get to single cycle execution.

Well, but the reason is that probably a single cycle full 128bit execution path was too costly (hardware-wise) back then.
And with that implementation, SSE was roughly equivalent to 3dNow! in performance.
I mean, of course we'd like to have very wide and low latency execution everywhere, but a CPU design is made mostly of tradeoffs (example, the P4 and its crippled FPU and complex integer execution units).

 

[SuperFly03]

--side stepping all Dell + AMD comments because they are irrelevant

What is amazing is that since 90nm we have gone from 1 core to 4 cores in 65nm when before we barely got 1 core on a cpu for what? 30 years (don't know exact timeline), and yet in the past 18 months we have gone from 1 to 4 cores? Ridiculous, and impressive.

I never had a problem with Prescotts heat level, my arguement was that between Prescott and AMD you could get the same performance with a cooler chip. So, if Intel does decide to keep Kentsfield at about the same TDP as a prescott (or slightly under) I am ok with that because it is a beast of a chip that really tears up any competition (as of now). It is all about cost/benefit, for a prescott there was little cost/benefit (lets not get into a hay day arguement about prescott) reletively, but Kentsfield shows some raw muscle. Personaly, I say just buy a solid aftermarket heatsink, slap it on and BAM you temps should drop into the 50's under load which is warm but nothing ridiculous.

 

[geralt]

Given the fact that Dell has over 50% Pentium 4 systems in their offering, repleacing them with Athlons is quite smart.

 

[geralt]

About TPD.
As far as I know it stands for Thermal Power Design, i.e. power dissipated by the processor. The power used by the processor is a sum of TPD and power needed for computing tasks.
TPD is always lower than power consumed.

Why everybody seems to refer TPD as power consumed by the processor ?

 

[SuperFly03]

About TPD.
As far as I know it stands for Thermal Power Design, i.e. power dissipated by the processor. The power used by the processor is a sum of TPD and power needed for computing tasks.
TPD is always lower than power consumed.

Why everybody seems to refer TPD as power consumed by the processor ?

basic Explination

"represents the maximum amount of power the thermal solution in a computer system is required to dissipate."

It's the number we are all concerned about because of cooling needs and also gives an indication of headroom (kinda of). (Edited due to research)

Intel is listing TDP numbers that are significantly lower than the actual maximum power draw of their CPUs. They are then relying on the fact that most applications barely use the CPU, assuming that it will remain idle most of the time. In the case that an application does max out the CPU for any period of time, Intel relies on their “Thermal Monitor” to automatically slow down the CPU when it becomes too hot to protect it from overheating.

AMD, on the other hand, lists TDP numbers that are significantly higher than the maximum power draw of their CPUs. They also have listed the SAME TDP for every desktop Athlon 64 so far, and I have little reason to believe that future Athlon 64s will have a higher listed TDP (at least for the near future). We still don't have definitive information about the exact power dissipation of each of the Athlon 64 processors, but it is clear that other than the fastest clock models, it is far below the 89W TDP cited by AMD. "

 

[ElMoIsEviL]

K8L better be bloody fast at a good price and they better hurry up, things aint lookin good

DELL just bought 2 million + Athlon
COVERTING 1/3 RD of its line to AMD

Intel is cutting jobs like by the 10's of thousands....

tell me....what kind of availability do you expect KENSFIELD to show...

Hey buddy... what did AMD do in 2002? That's right, they laid off 15% of there employees. It's called streamlining. Intel is undergoing major restructuring. It was a bloated ship and is being stirred into a different direction.
AMD did the same thing before and while launching there VERY successful K8 line.

Don't believe me?
CLICK ME!

In retrospect Intel has cut 10% of it's workforce. Far less then AMD's 15%. It's all relative. Intel Employs more people, but they're also selling off useless assets.

Do you honestly believe that 4x4 will sell that well? Other then enthusiasts I'm fairly certain that not too many people will pay up to $1000USD for a Dual Athlon64 FX CPU package (that's right you'll need to buy them in a dual pack), then they'll need a 4x4 motherboard which should cost more then current high end 975 boards (which will drop in price, but also seeing as the 4x4 boards will need MANY extra components) and also you'll need 4 STICKS of RAM! To run 4x4 (two for each processor) and to top it off.. it will be slower then Core 2 Quadro (Kentfield)... LMAO. (Each K8 core is slower then each Core 2 core and neither of the two are suffering from bandwidth issues).

Before making outlandish comments you'd be best to wait it out. Intel will dominate at least until K8L. After which the processor market should become quite competitive. Until then it's all Core 2... whether you like it or not.

Also we all know who will be posting some losses come next quarter and next year. If history is any indicator.. once a price war starts. AMD tend to post losses. CLICK ME!

 

[cxl]

About TPD.
As far as I know it stands for Thermal Power Design, i.e. power dissipated by the processor. The power used by the processor is a sum of TPD and power needed for computing tasks.
TPD is always lower than power consumed.

Why everybody seems to refer TPD as power consumed by the processor ?

Ahh, I understand that you are new to physics, so please let me explain a thing to you:

More than 99% of power that the computer consumes leaves the case as heat. There is no "power needed for computing tasks" that would not ended as heat.

(As for that miniscule rest, some power ends as light of LEDs, some ends as electric current of your ethernet card or VGA connection (and ends as heat in elsewhere), maybe you can have Wi-Fi card, in that case it ends as EM radiation etc... And yes, some power is used to open DVD drive tray .

 

[vikris]

Only patience is a path to a salvation!

I'm sure that my next upgrade will be when 100 core cpu's are out!

 

[npilier]

[quote="ElMoIsEviL]
Do you honestly believe that 4x4 will sell that well? Other then enthusiasts I'm fairly certain that not too many people will pay up to $1000USD for a Dual Athlon64 FX CPU package, [...] efore making outlandish comments you'd be best to wait it out.
[/quote]

I agree with you...Kentsfield will be a solid solution, but how could someone know that Kentsfield will be better than 4x4? There are not ES samples out yet anywhere so we may want to "wait and see." Also per the official review here at THG, Kentsfield will be priced at the same range as a 4x4 solution will be. Although I have to admit that Kentsfield will be a better upgrading path -at the moment- than any AMD solution.

Don't lose your time arguing with Lordpope, it's futile!

 

[Talon]

All I can say is like some other posters I'm sure glad I was waiting to buy C2D til there was better board availability. This just means as I save more $ and the new boards hit Kentsfield might be here and I'll just get quadro instead. Think of all the movies I can burn!! Backups and personal videos of course

 

[Wombat2]

I agree with you...Kentsfield will be a solid solution, but how could someone know that Kentsfield will be better than 4x4?

Because AMD can currently only make the K8 architecture i.e. X2s and Opterons. Do you seriously expect AMD to design a new wonder chip from the ground up and manufacture it within 3 months?! Thats simply impossible because it takes 2 years to do that. (Its going to happen ... eventually ... with the K8L launch in H2 2007).

The ONLY chips AMD has available to put into the 4X4 platform are re-badged Opterons whose performance we know already in detail. (Ive had them for a year for example.)

The marketer who came up with the 4x4 idea certainly deserves a pay rise. AMD really are fooling the unwashed masses

 

[JonathanDeane]

All I can say is like some other posters I'm sure glad I was waiting to buy C2D til there was better board availability. This just means as I save more $ and the new boards hit Kentsfield might be here and I'll just get quadro instead. Think of all the movies I can burn!! Backups and personal videos of course

couldnt agree more.

 

[gOJDO]

SSE was roughly equivalent to 3dNow! in performance.

3D Now! is not equivalent to SSE. It is much more closer to MMX than it is to SSE.

 

[gOJDO]

Imagine if Intel could get all four cores to work in series. 4x3ghz cores, 12GHZ! Wow, imagine the performance in single threaded apps. But yeah, go quad core. BTW, THG, could I perhaps take that kentsfield of your hands?

Yeap, can you imagine where I work we have >500 PCs, all >2GHz. Imagine 1THz with 512GB of RAM.

 

[mjp1618]

C'mon, we all know that Intel uses a different method (UL consumer goods spec) than AMD (maximum possible disspation: Vcore * Imax) so the TDPs cannot really be compared. The maximum power dissipation of a 65W Conroe is 87 W (Vcore * Imax) while the 65W X2 EE is actually 65W. It is also funny to note that Intel rates its Xeons, or at least did with the P4-based Xeons, by 90% of Vcore*Imax instead of the 75% as per the UL spec. So a 4x4 with AMD's EE processors can draw as much as 130W while the Kentsfield can draw 174W. No wonder during Tom's testing it was in the upper 60s like the hottest of the Smithfields.

It looks like it will be a long road to having more than 4 cores. Intel's problem will be keeping power dissipation low enough and AMD's problem will be to fit it all on one die. Maybe when Intel gets to 45nm we'll start to see rumblings of 8-core chips and when Intel hits 32nm and AMD hits 45nm we'll get them.

I thought Intel declared TDPs for product families, not individual products. So, just because a chip belongs to a family with 65W TDP does not mean that it's maximum power draw under normal working conditions (or whatever) is 65W.

It has been shown in several benchmarks (like X-bitlabs) that many 65W TDP CPUs from Intel consumes a lot less than 65W even under full load. So, it is not really 87W.

Contemporary Dual-Core Desktop Processors Shootout

 

[GloriosoSLB]

"and also you'll need 4 STICKS of RAM! To run 4x4 (two for each processor)"

Not really in fact even one should work (not recommended).

The 4x4 will look similar to this (Only one processor gets RAM):
http://www.tyan.com/products/html/tigerk8w.html
http://www.tyan.com/products/html/thunderh1000e.html
http://www.msi.com.tw/program/products/mainboard/mbd/pro_mbd_detail.php?UID=669

Not like you said that would look like this:
http://www.tyan.com/products/html/thundern3600b.html

The 4x4 will have to keep design simple, to keep it cheap.


If Intel came with quad core for 1000$ or AMD with two dual for 1000$, I think there will be no problem with either solution.

Power consuming should be low like Jumping Jack said. Intel had (still as) the 840EE and others that consume much more power than any of Intel or AMD future solution. So power consuming won’t be a problem.

Off topic: How much energy does the "recommended" toms PD805 OC uses?

 

[Wombat2]

Off topic: How much energy does the "recommended" toms PD805 OC uses?

500W

hehe ... the P4 arch sucked so much ass

 

[LordPope]

the DELL AMD comment was made because somone said " oh god KENSFIELD is so great AMD is in big trouble"

so i pointing out the the DELL deal to show...no AMD is not in trouble...

 

[SuperFly03]

Well AMD is in trouble of a sort, but they have been there before. Intel has really come back strong (stock withstanding) and with the recent job cuts, they are trimming themselves after a long hiring spree which was unnecessary. AMD is attempting to grow towards Intel's size by acquiring ATI for their chipsets (and GPUs), and for too long have been basking in the glory of the K8 (with good reason) but they kind of missed the boat on timing and process manufacturing.

They are falling behind and are going to have to work hard to regain their performance edge. I realize that the server market is an entirely diffrent story but if Intel comes back just as hard against that market as theyy have come back in the desktop segment, AMD could be in for a rough patch. Will they go bankrupt? of course not, but that doesn't mean they won't have to reset their priorities and really push some new products out the door.

I am trying to bridge the gap between OO AMD is screwed and Intel Rules to a more grounded sense of reality.

 

[ElMoIsEviL]

"and also you'll need 4 STICKS of RAM! To run 4x4 (two for each processor)"

Not really in fact even one should work (not recommended).

The 4x4 will look similar to this (Only one processor gets RAM):
http://www.tyan.com/products/html/tigerk8w.html
http://www.tyan.com/products/html/thunderh1000e.html
http://www.msi.com.tw/program/products/mainboard/mbd/pro_mbd_detail.php?UID=669

Not like you said that would look like this:
http://www.tyan.com/products/html/thundern3600b.html

The 4x4 will have to keep design simple, to keep it cheap.


If Intel came with quad core for 1000$ or AMD with two dual for 1000$, I think there will be no problem with either solution.

Power consuming should be low like Jumping Jack said. Intel had (still as) the 840EE and others that consume much more power than any of Intel or AMD future solution. So power consuming won’t be a problem.

Off topic: How much energy does the "recommended" toms PD805 OC uses?

I wish that were true.. but AMD is going for dedicated memory bandwidth per processors... check out there official diagram bellow. See you'll need 4 sticks of RAM for Dual Channel per processor (Quad Channel effective). You could get one stick per channel but then you'd severly be limiting memory bandwidth.

 

[clue69less]

What is amazing is that since 90nm we have gone from 1 core to 4 cores in 65nm when before we barely got 1 core on a cpu for what? 30 years (don't know exact timeline), and yet in the past 18 months we have gone from 1 to 4 cores? Ridiculous, and impressive.

I'm all over that. I honestly feel that '06 will be regarded as a major turning point, where hardware got so far ahead of OS and software for the majority of users, it's basically an endpoint of sorts.

I never had a problem with Prescotts heat level, my arguement was that between Prescott and AMD you could get the same performance with a cooler chip. So, if Intel does decide to keep Kentsfield at about the same TDP as a prescott (or slightly under) I am ok with that because it is a beast of a chip that really tears up any competition (as of now). It is all about cost/benefit, for a prescott there was little cost/benefit (lets not get into a hay day arguement about prescott) reletively, but Kentsfield shows some raw muscle. Personaly, I say just buy a solid aftermarket heatsink, slap it on and BAM you temps should drop into the 50's under load which is warm but nothing ridiculous.

C2D basically is a predictor in my mind. It sets a power vs. performance benchmark and it's only a matter of time before C2Q or whatever will slide into the same ballpark. Yes, raw muscle is a good way of putting it. Next up, refined, slimmed down muscle!