if a yorkfield is two wolfdales...

raging bull · 06-30-2008, 10:37 PM

then why wouldnt a qxxx overclock just as well as an e8xx?

ctal · 07-01-2008, 07:55 AM

Quote:

Originally Posted by raging bull

then why wouldnt a qxxx overclock just as well as an e8xx?

Because with more transistors in the package, there is more heat. Higher temperatures = slower performance of individual transistors (that's inherent in their physics) = longer delays for signals to propagate through the internal circuit paths = slower overall performance and less overclocking potential.

It's not just overclocking that is reduced. Stock speeds are lower as well, unless you are willing to pay a huge cost premium for the fastest quads.

Regards,
-- Al

**Gorganzola** · 07-01-2008, 10:49 AM

Beyond the heat, you also have additional overhead for the Cores talking to each other. It is not a 1:1 correlation with additional cores resulting in equivalent increases in performance.

This is were kONGO jumps in with the link to the wiki on this theory.

XJ. · 07-01-2008, 01:03 PM

Well...
1) Now you have two dual cores, each with it's own limitations - so you can only overclock as far as the weakest core will go.
2) The quad core puts allot more strain on the power regulator and possibly other circuits on the mobo making it more likely that you will hit some sort of limit earlier.
3) Quads burden the FSB more, not sure exactly why because I haven't studied the specifications and design considerations - but this is clearly the case from my own and others overclocking experiences.

ctal · 07-01-2008, 01:48 PM

To add a more quantitative perspective to my previous post, the Thermal Design Power (which is what Intel specifies as the target dissipation upon which thermal solutions should be based) is 65 watts for the E8xxx's, while for the Q9xxx's it is 95 watts.

These numbers can be higher under conditions of prolonged heavy cpu utilization, and will also increase if the clock rate is raised and/or if the supply voltage is raised (which is commonly done to optimize overclocking).

The answers provided by Gorgo and XJ are certainly very valid, but I think it is the thermal difference that most directly addresses your question. Basically, as temperatures go up, transistors (and integrated circuits comprised of transistors) slow down.

Regards,
-- Al

06-30-2008, 10:37 PM	#1
raging bull Registered User Join Date: Nov 2003 Location: north carolina Posts: 285	if a yorkfield is two wolfdales... then why wouldnt a qxxx overclock just as well as an e8xx? __________________ P5KdeluxeWiFi/E6750(2.66-NoOC)C2D/2X1G-Mushkin eXtremePerformanceDDR2-1066(PC2-8500)/Two320GB SeagateBarracuda7200.10s/Lite-On20X/Pioneer18X/AsusX1950Pro-256/X-Fi XtremeGamer/Corsair520HX/LianLiA16B/VistaBusiness P4P800e/2.4c@3.1(1.79v);2x1G-OCZ5001024ELGEGXT(3-4-3-8-1T) WD36/WD200;Memorex16x/Lite-on52x;eVGA6800GS256AGP; SBAudigy2;Enermax 465 T61/C2D-T7300(2.0G)/2GB-PC2-5300/DDR2-667/nVidia Quadro NVS140/100GB-HD-7200rpm/DVD-recordable 8X-DL/Bluetooth/VistaBusiness
(Offline)

07-01-2008, 10:49 AM	#3
Gorganzola Stuck in 3D Join Date: May 2001 Location: Hangin' with the fruits Posts: 9,343	Re: if a yorkfield is two wolfdales... Beyond the heat, you also have additional overhead for the Cores talking to each other. It is not a 1:1 correlation with additional cores resulting in equivalent increases in performance. This is were kONGO jumps in with the link to the wiki on this theory. __________________ TTFN. I wasn't asleep at the switch, I was drunk. -- Homer J. Simpson Q. How many dull people does it take to change a lightbulb? A. One. A very useful tool on these forums: You can Meebo in public.
(Offline)

07-01-2008, 01:03 PM	#4
XJ. ... Join Date: Apr 2002 Location: NH, USA Posts: 6,004	Re: if a yorkfield is two wolfdales... Well... 1) Now you have two dual cores, each with it's own limitations - so you can only overclock as far as the weakest core will go. 2) The quad core puts allot more strain on the power regulator and possibly other circuits on the mobo making it more likely that you will hit some sort of limit earlier. 3) Quads burden the FSB more, not sure exactly why because I haven't studied the specifications and design considerations - but this is clearly the case from my own and others overclocking experiences.
(Offline)

07-01-2008, 01:48 PM	#5
ctal Eschews Obfuscation Join Date: Sep 2003 Location: Connecticut, USA Posts: 1,240	Re: if a yorkfield is two wolfdales... To add a more quantitative perspective to my previous post, the Thermal Design Power (which is what Intel specifies as the target dissipation upon which thermal solutions should be based) is 65 watts for the E8xxx's, while for the Q9xxx's it is 95 watts. These numbers can be higher under conditions of prolonged heavy cpu utilization, and will also increase if the clock rate is raised and/or if the supply voltage is raised (which is commonly done to optimize overclocking). The answers provided by Gorgo and XJ are certainly very valid, but I think it is the thermal difference that most directly addresses your question. Basically, as temperatures go up, transistors (and integrated circuits comprised of transistors) slow down. Regards, -- Al __________________ "I didn't say I didn't say it. I said that I didn't say that I said it. I want to make that very clear." -- George Romney, in 1968, while campaigning for the Republican nomination for President of the United States.
(Offline)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode