Core 2 Extreme QX9650 To 4GHz (stable)

14789 Views 31 Replies 13 Participants Last post by 10acjed, Oct 31, 2008

Amajed

Discussion Starter · Oct 31, 2008

11

Hi every one

i'm glad to be here in this forum and I learned a lot from you guys

so after reading to many threads I decide to OC my friend's CPU QX9650 to 4GHz which I reached today with fully stabled system

heres the SS for the OC :

http://aam.1407.googlepages.com/OverClock.jpg

As u can see the Prime95 started at 13:13 and the SS was taken 14:15 (more than one hour , and its still running now at this time)

OK heres the thing

the voltage for the CPU is 1.5v "I know most of u will say too much" but this is the only voltage I can fully test it without crashing.

I raised the voltage 0.05 every crash from 1.3 till I reached 1.5 and finally stale system

also this article says its OK to go for 1.5 for a daily usage setup:
http://www.nextlevelhardware.com/cpu/qx9650/index.html

Quote:

We are pushing 1.50 max vcore into this QX, and using my judgement I feel this would be ok for a daily usage setup.

(Correct me if I'm wrong plz)

so thats the whole story ppl

I hope I can get a stable 4GHz with less voltage than what I did back there its my friends CPU not mine thou

BTW the temp is too High "*He is now using the Intel QX9650 CPU Cooler." and I recommended this one:
Thermalright TRUE Copper

and I think he will get it soon

, if there is any better heat sink for this CPU now is the chance to correct it

System spec:
Intel Core 2 Extreme QX9650 @ 4GHz 1780FSP 1.5vcore
EVGA 132-CK-NF78-A1 nForce 780i
Kingston HyperX KHX9600D2K2/2G @ 5-5-5-15 1066MHz 2.2v
NVIDIA GeForce 9800 GX2 stock settings

Thanks for reading this

its open for Comments and recommendations

UPDATE**:
I stopped the Prime95 TEST after 4 Hours with no errors, here is pic's for before and after I stopped the TEST,

Before I stopped the TEST:
http://aam.1407.googlepages.com/Before.jpg

After I stopped the TEST:
http://aam.1407.googlepages.com/After.jpg

See less See more

1 - 20 of 32 Posts

robbo2

· #2 · Oct 31, 2008

The temps are quite high an i though 1.5 would be alot for a 45mn?

Lukeatluke

· #3 · Oct 31, 2008

TRUE is the best.And voltage is to high for 45nm, max is 1.3625V

See less See more

Kyle659

· #4 · Oct 31, 2008

Volts seems high to me, max for 45nm is around 1.368.

Psycho666

· #5 · Oct 31, 2008

drop the voltage to 1.3625 and you`ll be safe...now you`re just killing the cpu

See less See more

10acjed

· #6 · Oct 31, 2008

voltage set in bios for 45nm chips should be no more than 1.3625, even if it reads lower in windows. This chip wont last long at that high of a voltage and temp.. The stock heatsink is almost useless, a TRUE will bring temps down to the 55-60c area which is better...

Also this only shows orthos running for about 9 minutes, thats not stable, and within a half hour the temps will probly exceed 72c...

Id say get a better cooler first, then put the voltage at 1.360 in bios, and push the FSB voltage to 1.35-1.4 and then see how high it will go... once unstable you have probably found your limit...

Amajed

Discussion Starter · #7 · Oct 31, 2008

3

Thank you for these very fast reply

Unfortunately the voltage can NOT be reduced anymore , when its under 1.5 the system will crash "BLUE Screen"

if there is anyway to prevent the crash with less voltage I would be happy to try it

Just to make things clear I tested it with 1.368 voltage and the BLUE Screen appears after I started the Prime95 test 1min passed only .

TEST completed about 2 Hours and its still running

See less See more

Takendown2

· #8 · Oct 31, 2008

if ur doing a suicide run then 1.5V is ok but for 24/7 i would not exceed 1.3625V

10acjed

· #9 · Oct 31, 2008

5

Quote:

Originally Posted by Amajed

View Post

Thank you for these very fast reply

if there is anyway to prevent the crash with less voltage I would be happy to try it

TEST completed about 2 Hours and its still running

Its not "safe" to run at that voltage... Specially the stress test... But, its your chip...

See less See more

Amajed

Discussion Starter · #10 · Oct 31, 2008

6

Quote:

Originally Posted by 10acjed

View Post

voltage set in bios for 45nm chips should be no more than 1.3625, even if it reads lower in windows. This chip wont last long at that high of a voltage and temp.. The stock heatsink is almost useless, a TRUE will bring temps down to the 55-60c area which is better...

Also this only shows orthos running for about 9 minutes, thats not stable, and within a half hour the temps will probly exceed 72c...

Id say get a better cooler first, then put the voltage at 1.360 in bios, and push the FSB voltage to 1.35-1.4 and then see how high it will go... once unstable you have probably found your limit...

Thank you for ur reply

Blue:
when u say chip wont last too long , How long do u mean by that ? "considering the voltage not the temp"

Red:
I really didn't understand most of it

what is the "orthos" ? where did u find 9 minutes exactly ? also the prime95 test is now running about 2 hours and the temps are the same as it shown on the SS.

Green:
Too bad I don't think I can test it again for many reasons

so this is my last day of testing this CPU

if there is any changes I can do right now I would be happy to try it

EDIT:
PS, My friend want the CPU clock 4GHz at least (thats why he bought this chip)

See less See more

Lukeatluke

· #11 · Oct 31, 2008

As much as i seen even q9550 exceeded 4.0 under 1.3625V and that kind of voltage for even better(read=more expensive) chip?

See less See more

Matt*S.

· #12 · Oct 31, 2008

I would say that he can't notice the difference between 4.0GHz and 3.8 which you should be able to get stable with significantly less voltage. You can try playing with the reference voltages if his board has the option for it, or you can increase FSB Termination Voltage a little bit, I wouldn't go over 1.35v.

As it is, @ 4.0GHz he's going to be looking for another Processor WAY sooner then he should have to. Currently we don't necessarily know the high voltage effects on 45nm, but I have heard of people making a nice conversation piece, and then looking for a new processor @ 1.5v.

Wattes

· #13 · Oct 31, 2008

- 1.5v is way too much for 45nm. I know and have seen people on this forum that had there 45nm's die in weeks with "just" 1.45v.

- I don't think any 45nm needs more then 1.4v to get to 4Ghz. Most people make it with just 1.35v (in fact, my Q9650 runs @ 4Ghz now under just 1.31v and I haven't tweaked a thing yet; but it is a different model/stepping, so you can't compare them 1:1).

- Like 10acjed said, I would look at other settings first and definatly never go over 1.45v unless you want to do suicide runs with 5Ghz+

- For the best stability keep them under 60C; so yeah a better cooler will help.

- Mind posting all the settings in your BIOS (voltages/timings/mhz), so we can help you better?

<3

See less See more

burning-skies

· #14 · Oct 31, 2008

4

Quote:

Originally Posted by Amajed

View Post

Hi every one

i'm glad to be here in this forum and I learned a lot from you guys

OK heres the thing

the voltage for the CPU is 1.5v "I know most of u will say too much" but this is the only voltage I can fully test it without crashing.

also this article says its OK to go for 1.5 for a daily usage setup:
http://www.nextlevelhardware.com/cpu/qx9650/index.html

(Correct me if I'm wrong plz)

did you not read this little bit in the article....

Quote:

Not only were we able to achieve a 24/7 overclock on this CPU for daily usage with stock voltage. But bumping up voltage to 1.65 vcore on this processor will allow you to achieve a staggering 57% overclock and easily break into the 9 second barrier using SuperPi 1m as a benchmark. **Please do not attempt to push high voltage into your processor as it can be detrimental to longevity especially without ample cooling. Maximum recommended voltage is listed at 1.3625 for this processor.**

your guna see alot of money go down the drain very soon!

See less See more

ericeod

· #15 · Oct 31, 2008

I've seen a 45nm CPU die in just under 2 weeks running at 1.45v vcore (Tricky's E8400). I think the best thing to do is to reduce the clock speed and voltage. Also, what FSB Termination (VTT) are you running? It needs to stay below 1.4v.

The QX9650 is an older core which is either a C0 or C1. They dont OC as well as the E0 core and the older cores have a max operating temp of 64.5Â°C while the E0 have a max operating temp of 71.4Â°C. If your friend can return it and get a Q9650 (E0), he shouls be able to reach 4.0GHz+.

Mmansueto

· #16 · Oct 31, 2008

i had a QX9650 and with 1.38 volts i was able to stay stable at 4.2ghz. you should have no problem getting 4.0 with less than 1.5volts. There is some setting that you are overlooking. Please post all your BIOS settings. Also, what MB does he have?

Amajed

Discussion Starter · #17 · Oct 31, 2008

3

OK here we are with the BIOS settings

First I wanted to thank you all for these replys

1- System Clocks:
http://aam.1407.googlepages.com/DSC02706.jpg

2- FSB & Memory Config:
http://aam.1407.googlepages.com/DSC02707.jpg

3- Memory Timing Setting:
http://aam.1407.googlepages.com/DSC02708.jpg

4- CPU Configuration:
http://aam.1407.googlepages.com/DSC02709.jpg

5- System Voltages:
http://aam.1407.googlepages.com/DSC02710.jpg

Please see all these pic's before comment

all these pic's are very small (38KB Only)

EDIT:
the MB name is :
EVGA 132-CK-NF78-A1 LGA 775 NVIDIA nForce 780i SLI

BIOS ver. P08

See less See more

Mmansueto

· #18 · Oct 31, 2008

What board is this?

EDIT. nevermind I just saw what you had.

Matt*S.

· #19 · Oct 31, 2008

I can't guarentee this is your problem, but turn off all of the Spread Spectrum stuff and try again, usually see more stable OC's.

Mmansueto

· #20 · Oct 31, 2008

First I would lower the FSB to 400, and increase the multi to 10. then I would lower the vcore to around 1.38125 or so. Then I'd try to play with the other voltages, like NB, etc.

1 - 20 of 32 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.