Xtreme Low Q6600 G0 Temp on Air !!! (Lots of Pics)

[XtremeTiramisu]

Just to add to your info above, they've changed the Xeon parts to some new specs recently...

A X3220 G-0 has (equiv to Q6600)
Thermal Design Power: 100W
Thermal Specification: 85C

A X3210 G-0 has (equiv to non-existent Q6400 @2.13GHz, 8x multiplier)
Thermal Design Power: 100W
Thermal Specification: 85C

...so it looks like the server parts have more headroom.

Nice info.
But even with greater delta for more headroom doesn't necessary means decent overclock-ability. In the end, it may ended up with a similar OC but at a higher temp and still within the thermal spec. The additional +5W consumption does makes sense to also increase the thermal spec up to 85C.
It's still a gamble and luck of draw but I'd personally go with a lower TDP and relatively lower thermal spec CPU to OC than higher TDP + higher Thermal Spec CPU. This is how my logic and experience has given me.

The only way to really draw some clear statements is by accumulating enough OC database for the new X3220/3210 and compare them to the G0 Q6600

[Spanki]

Yeah, noone's seeing G0 parts yet for those, so it might be a while before we know how they clock.

The interesting thing about the X3210 is that it can be found for ~$225 today (though you'll surely get a B3 stepping).

I'll be interested to see what kind of VIDs the G0 stepping Xeons end up with. You could make an arguement for either better or worse binning - but we won't know until we see some in action.

[idiotekniQues]

Here are the results of Speedfan VS. CoreTemp. Disregard the Freq in Core Temp, its wrong. After seeing this, I dropped the OC back to 3.2ghz. So its running at 44c idle at 3.2ghz. I need a better cooler.

question. did you stress test the chip and for how long?

i am taking results from some forums and giving them to a guy at anandtech who is putting them all together.

[orion23]

I just realized that I applied the thermal paste the wrong way.

I did it the same way I did my Dual Core CPU's and now I see that it is supposed to be the complete opposite g

Oh well, I'm getting good temperatures I think.

35c Idle or so.

Maybe if I decide to lap my cooler, then I'll reapply the paste the right way!

[XtremeTiramisu]

FSB 470 * 8 @ 3760MHz 8Hrs+ LargeFFT Prime Stable - Final stable OC in the summer season.


As per requested, here's the 8Hrs+ LargeFFT Prime Stable screenshot at your disposal
LargeFFT Stress test is known to output the maximum heat, power consumption and stress to the entire system.
P5K-Dlx, Q6600 G0 and Ballistix Tracers all sustained the long 8Hrs+ torture at the given voltage : clock speed
For reference purposes, I have also attached the 8Hrs+ SmallFFT Prime stable screenshot below for you to see "a noticeable difference in the LOAD Vcore".
It appears that LargeFFT does actually puts out more stress so the Vcore further drops down to 1.4720v from 1.4800 OCCASIONALLY.
Whereas in SmallFFT, the load Vcore always remained at 1.4800 flat throughout the 8Hrs+ Prime.
So both results yields that the system is rock stable with its current OC settings and I must say it again - What a FANTASTIC OC result on air






Finally the moment you all have being waiting for: Complete Asus P5K-Deluxe BIOS Configuration. FYI


Please feel free to use it for your reference. The configurations shown above may not be 100% identical for all cases.
You're welcome

JumperFree Configuration Settings
AI Overclocking: Manual
CPU Ratio Control: Manual
- Ratio CMOS Setting: 8
FSB Strap to Northbridge: AUTO
FSB Frequency: 470
PCI-E Frequency: 100
DRAM Frequency: DDR2-1175
DRAM Command Rate: 2T
DRAM Timing Control: Manual
CAS# Latency: 5
RAS# to CAS# Delay: 5
RAS# Precharge: 4
RAS# Activate to Precharge: 7
TWR: 3
TRFC: 30
TWTR: 3
TRRD: 3
TRTP: 6
DRAM Static Read Control: ENABLE

Transaction Booster: AUTO
Boost Level: (only displays when Transaction Booster is set to Disable or Enabled)
Clock Over-Charging Mode: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

CPU Voltage: 1.5250
CPU Voltage Reference: x0.63
CPU Voltage Damper: Enabled
CPU PLL Voltage: 1.7
DRAM Voltage: 2.2
FSB Termination Voltage: 1.40
North Bridge Voltage: 1.55
North Bridge Voltage Reference: x0.67
South Bridge Voltage: 1.2

Advance CPU Settings
CPU Ratio Control: Manual
- Ratio CMOS Setting: 8
C1E Suppport: Disabaled
Max CPUID Value Limit: Disabaled
Vanderpool Technology: Disabaled
CPU TM Function: Disabaled
Execute Disable Bit: Enabled
PECI: Enabled

USB Configuration
USB Functions: Enabled
Legacy USB Support: Disabled (need to enable it to detect USB flash drives)

[XtremeTiramisu]

I just realized that I applied the thermal paste the wrong way.

I did it the same way I did my Dual Core CPU's and now I see that it is supposed to be the complete opposite g

Oh well, I'm getting good temperatures I think.

35c Idle or so.

Maybe if I decide to lap my cooler, then I'll reapply the paste the right way!

Well I guess you probably have seen the quad core application method illustrations I posted in the previous page so now you would know

Oh and look what I've for you above this post

[idiotekniQues]

I just realized that I applied the thermal paste the wrong way.

I did it the same way I did my Dual Core CPU's and now I see that it is supposed to be the complete opposite g

Oh well, I'm getting good temperatures I think.

35c Idle or so.

Maybe if I decide to lap my cooler, then I'll reapply the paste the right way!

hey, is your chip stress-tested?

let me know how much.

[orion23]

Well I guess you probably have seen the quad core application method illustrations I posted in the previous page so now you would know

Oh and look what I've for you above this post

Awesome! I've been waiting for that!

Hopefully I'll have extra time this weekend to play with the 8X multiplier instead of the default 9.

Thanks for posting the settings!

hey, is your chip stress-tested?

let me know how much.

Well, @ 3.7ghz, using 1.5V, it idles anywhere from 35c to 40c.

I also tried 3.9ghz with 1.6125V and it would idle @ 38c - 45c which I thought was stranged!

[treatmentx]

I'm hitting a weird wall, I need help =(

I'm on a p5n32-E SLI, G0 Q6600. I updated to the latest 1203 BIOS, and now if I hit anything over 360FSB. Sometimes it would make it to windows, other times it wouldn't, it would just make a high pitched sustained beep.

I'm currently stable at 3.2GHZ (355MHZ), 30/29/29/29C idle temps
Mem : 2.2v
CPU VTT: 1.55v
1.2VHT : 1.40v
NB : 1.50v
SB : 1.55v

any suggestions?

[fallwind]

965 boards are not the greatest for overclocking quads. You'll normally run into a FSB hole between 360 and 400MHz. Try 8x401 which kicks in 1333 FSB strap(400 and below is on the 1066 strap). If it works then you can proceed up from there but don't expect to get much farther.

[treatmentx]

I see... did you mean 965 or did you mean 680i? The P5N32-E board is a 680i board.

[Charloz24]

Excellent thread XtremeTiramisu! Glad to see some really good results comming from Canada! I finally pulled the trigger on a Q6600 G0 and P5K premium on my watercooling setup! So my credit card is loaded because of you!

[jeffyjaixx]

I'm hitting a weird wall, I need help =(

I'm on a p5n32-E SLI, G0 Q6600. I updated to the latest 1203 BIOS, and now if I hit anything over 360FSB. Sometimes it would make it to windows, other times it wouldn't, it would just make a high pitched sustained beep.

I'm currently stable at 3.2GHZ (355MHZ), 30/29/29/29C idle temps
Mem : 2.2v
CPU VTT: 1.55v
1.2VHT : 1.40v
NB : 1.50v
SB : 1.55v

any suggestions?

Don't worry ... I'm hitting that 360 FSB wall here as well on a 965P S3.

[Clint]

FSB 470 * 8 @ 3760MHz 8Hrs+ LargeFFT Prime Stable - Final stable OC in the summer season.


As per requested, here's the 8Hrs+ LargeFFT Prime Stable screenshot at your disposal
LargeFFT Stress test is known to output the maximum heat, power consumption and stress to the entire system.
P5K-Dlx, Q6600 G0 and Ballistix Tracers all sustained the long 8Hrs+ torture at the given voltage : clock speed
For reference purposes, I have also attached the 8Hrs+ SmallFFT Prime stable screenshot below for you to see "a noticeable difference in the LOAD Vcore".
It appears that LargeFFT does actually puts out more stress so the Vcore further drops down to 1.4720v from 1.4800 OCCASIONALLY.
Whereas in SmallFFT, the load Vcore always remained at 1.4800 flat throughout the 8Hrs+ Prime.
So both results yields that the system is rock stable with its current OC settings and I must say it again - What a FANTASTIC OC result on air






Finally the moment you all have being waiting for: Complete Asus P5K-Deluxe BIOS Configuration. FYI


Please feel free to use it for your reference. The configurations shown above may not be 100% identical for all cases.
You're welcome

Execute Disable Bit: Enabled

...why in heavens do you have that enabled??

Turn that off,..it is not doing you any good..on the contrary.

May I guess that you still have DEP enabled as well..?

[XtremeTiramisu]

Execute Disable Bit: Enabled

...why in heavens do you have that enabled??

Turn that off,..it is not doing you any good..on the contrary.

May I guess that you still have DEP enabled as well..?

1. Please educate me why on earth I should not have Execute Disable Bit Enabled?

2. Please elaborate what's DEP

Thanks.

[unclewebb]

Execute Disable Bit: Enabled

...why in heavens do you have that enabled?

Having that Enabled costs nothing in terms of performance and it will alert you when running some poorly coded or malicious software that is trying to access memory it shouldn't be. It won't limit your overclock in any way.

http://www3.intel.com/cd/ids/develop...eng/149308.htm

[XtremeTiramisu]

Having that Enabled costs nothing in terms of performance and it will alert you when running some poorly coded or malicious software that is trying to access memory it shouldn't be. It won't limit your overclock in any way.

http://www3.intel.com/cd/ids/develop...eng/149308.htm

Exactly what I've in mind too unclewebb, thanks for the explaination.
I really don't know what Clint was thinking.......

[[XC] DragonOrta]

I'm hitting a weird wall, I need help =(

I'm on a p5n32-E SLI, G0 Q6600. I updated to the latest 1203 BIOS, and now if I hit anything over 360FSB. Sometimes it would make it to windows, other times it wouldn't, it would just make a high pitched sustained beep.

I'm currently stable at 3.2GHZ (355MHZ), 30/29/29/29C idle temps
Mem : 2.2v
CPU VTT: 1.55v
1.2VHT : 1.40v
NB : 1.50v
SB : 1.55v

any suggestions?

That's a normal wall for your board. If you feel like doing some voltmodding, Kunaak has a thread where he did the GTL REF mod, and got his quad to 450fsb.

[eva2000]

Yeah, noone's seeing G0 parts yet for those, so it might be a while before we know how they clock.

The interesting thing about the X3210 is that it can be found for ~$225 today (though you'll surely get a B3 stepping).

I'll be interested to see what kind of VIDs the G0 stepping Xeons end up with. You could make an arguement for either better or worse binning - but we won't know until we see some in action.

X3230 G0 ES stock VID = 1.275v and AUTO vcore in bios confirms it at 1.248v real vcore when AUTO vcore is set. This is same VID as my Q6600 B3 ES cpu as well which managed on water cooling to be Wprime 32M + 1024M passable @3920Mhz at 1.616v in <14C ambients.

Room temp = 22.4C

Idle


Load


Intel Xeon X32xx Quad Core G0 stepping:

• Intel Xeon X3230 2.66Ghz - SLACS - G0 - 100W / TDP 85C - 1.100-1.372V
• Intel Xeon X3220 2.40Ghz - SLACT - G0 - 100W / TDP 85C - 1.100-1.372V
• Intel Xeon X3210 2.13Ghz - SLACU - G0 - 100W / TDP 85C - 1.100-1.372V

Intel Xeon X32xx Quad Core B3 stepping:

• Intel Xeon X3220 2.40Ghz - SL9UP - B3 - 105W / TDP 85C - 1.100-1.372V
• Intel Xeon X3210 2.13Ghz - SL9UQ - B3 - 105W / TDP 62C - 1.100-1.372V

Intel Core 2 Quad line:

• Intel Core 2 Quad Q6600 2.40Ghz - SL9UM - B3 - 105W / TDP 62.2C - 1.100-1.372V
• Intel Core 2 Quad Q6600 2.40Ghz - SLACR - G0 - 95W / TDP 71C - 1.100-1.372V
• Intel Core 2 Quad Q6700 2.66Ghz - SL9CQ - G0 - 95W / TDP 71C - 1.100-1.372V

[jkresh]

eva2000 any idea as to when the blitz formula will be available in the US?

[Spanki]

Intel Xeon X32xx Quad Core G0 stepping:

• Intel Xeon X3230 2.66Ghz - SLACS - G0 - 100W / TDP 85C - 1.100-1.372V
• Intel Xeon X3220 2.40Ghz - SLACT - G0 - 100W / TDP 85C - 1.100-1.372V
• Intel Xeon X3210 2.13Ghz - SLACS - G0 - 100W / TDP 85C - 1.100-1.372V

Slight typo there... the X3210 should be SLACU.

Thanks for the other info, but as we all know, ES != retail .

[Clint]

Having that Enabled costs nothing in terms of performance and it will alert you when running some poorly coded or malicious software that is trying to access memory it shouldn't be. It won't limit your overclock in any way.

http://www3.intel.com/cd/ids/develop...eng/149308.htm

Exactly what I've in mind too unclewebb, thanks for the explaination.
I really don't know what Clint was thinking.......

First of all, posting a lnk to an overview of an function with only a few lines of describing the intention of said function is not by any means considered a "proof" or valid argument...not by me anyway.
(it says nothing of the "cost" now, does it?)

Execute Disable Bit and Data Execution Prevention (DEP) are intimately close to each other, they WILL lower your systems efficiency due to the characteristics of "things".

NOTE: I write "characteristics of things" only because I've already promised certain mods here not to be so harsh on the "less knowledgeable" here.

Please please go out there and gather some info, and prefereably a more indepth one than that one you gave.

Anyway, I would disable Execute Disable Bit in BIOS

and DEP with this: (from Start Menu/Run)

bcdedit.exe /set {current} nx AlwaysOff

This WILL have an positive effect, trust me.

[ss256]

I just got my Q6600 from NCIX, same exact Batch #: L723A765 as XtremeTiramisu. It will be a week until I can get my D-TEK Fuzion on it. I'll keep you updated.

[unclewebb]

This WILL have an positive effect, trust me.

I posted that quick link to the Intel documentation so you or others could learn more about what it does and then you can make up your own mind whether you want it on or off.

I'm always interested in learning Clint. Do you know what type of benchmarks show how having the NX bit enabled decreases performance. I've never found any measurable difference in performance during testing but maybe there are programs out there where a performance hit can be measured.

The NX bit enabled has notified me of software that was trying to execute code in an area of memory that was reserved for data. A properly written program does not do that. The only time having the NX bit enabled will stop a program from running is if it is a malicious piece of code executing where it shouldn't be or it's a poorly written piece of code that is likely trying to access an uninitialized null pointer. I don't want either type of code running on my system so I have it enabled.

If you would have made it to page 2 of that link I posted then you would have come across this page which I think is a very thorough document about the NX bit:
http://www3.intel.com/cd/ids/develop...eng/149307.htm

[Clint]

I posted that quick link to the Intel documentation so you or others could learn more about what it does and then you can make up your own mind whether you want it on or off.

I'm always interested in learning Clint. Do you know what type of benchmarks show how having the NX bit enabled decreases performance. I've never found any measurable difference in performance during testing but maybe there are programs out there where a performance hit can be measured.

The NX bit enabled has notified me of software that was trying to execute code in an area of memory that was reserved for data. A properly written program does not do that. The only time having the NX bit enabled will stop a program from running is if it is a malicious piece of code executing where it shouldn't be or it's a poorly written piece of code that is likely trying to access an uninitialized null pointer. I don't want either type of code running on my system so I have it enabled.

If you would have made it to page 2 of that link I posted then you would have come across this page which I think is a very thorough document about the NX bit:
http://www3.intel.com/cd/ids/develop...eng/149307.htm

You are completely missing my point here, I still don't see any verification that it doesn't "cost" anything performancewise...the pdf tells nothing about the performance hit.

Everything comes at an price you know, and DEP for once is very well known to generate more problems than resolving them...

I also thought XtremeTiramisu was more inclined to getting the most out of his OC due to the vast amount of post and with the frequency he post them telling us about his fantastic chip...

But Ok, if you wanna go out on the net between the benches and posting of results chasing some pron...well by all means, keep those things.

To the rest:
Try this:

1.Do a fresh upstart of your box, and wait 5 min. (Vista related)
2.scribble down your memoryusage, then bench something

3a. In Vista use bcdedit.exe /set {current} nx AlwaysOff
3b. In XP add "/noexecute=alwaysoff" to your boot.ini (without ")

4.reboot and disable Execute Disable Bit in BIOS

5.fire up the OS and wait 5min (Vista related)
6. scribble down your memoryusage, then bench something

..should be an eyeopener, taken honest reflection is given..(that is, leaving the ePenis out of the equation..hehe )