• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

IP35-PRO and Q6600 help.

C

Copyright

Supreme [H]ardness
Joined
Nov 17, 2000
Messages
6,004
Picked up a used Q6600 and it seems to be a decent chip but cant seem to get 3.6ghz stable even wtih as much as 1.5V in windows (1.53 in bios). I can get about 20 min of prime stable and make it through Vantage mark with no problem but just cant get the 4 hours of prime stable I want. How high can I safely take the voltage? My max temps at 1.5V were 61C and the PWM was 74C. I do have have the bolt through mod. The Chip is lapped and so is the heatsink using MX-2 paste and the washer mod for more tension. Temps seem really good as I only hit 47C after an hour of orthos at stock clock. I did manage to get 8x425 stable for 4 hours @ 1.45V in bios and my PWM was 67-70C the whole time. Should I still keep aiming for 9x400 or maybe 8x450? Will it be worth all the extra voltage/heat I will have to throw at it? Thanks for any input or screen shots of your settings!
 
What are your ambient temps ? I have the same exact setup with bolt modded ip35 pro. Although I have Watercooling now, I hit 3.6ghz stable on my Tuniq Tower with similar temps to yours on air. Below is my max overclock of 3.8ghz that ran stable up to 2 hours, but you can try similar bios settings as I found these to be optimal. CPU vcore 1.50 idle 1.44 load(Vcore 1.53 in bios), MCH 1.41v, DDR2 2.20v , ICHIO 1.62v, ICH 1.09v, CPU VTT 1.23v, ambient 29C. My bios settings in the screenshot to the right under CPUID:
f_38StableVism_9bf94dc.jpg
 
If these settings don't work out for you granted all other parts are in order (PSU, good ram, etc) it could very well just be your Q6600 just not being a good overclocker.

edit also set GTLREF to 63%.
 
I'm running at Ip35 Pro w/ a Q6700 @ 3.6ghz as well. I've run all the combinations, but at the moment I'm running 360 x 10 w/ Vcore set at 1.545 in the bios (1.50-1.51 actual), 1.48nb, 1.42vtt, ram at 1080mhz. W/ 400 x 9 and 450 x 8 it's about the same settings ...maybe a little higher mch at 450. 3.8 is possible above 1.5 but that's not exactly safe water or no water.

On air you'll probably have to do the GTLREF tweaking like kadozer said. For some reason, when I went to water I stopped having to mess w/ that.

Also I'm using BIOS 16 beta 4.
 
What is GTLREF? the only thing I tried with that 69% and 70%... I didnt move the ICH10 at all. These other settings might explain why I the extra CPU voltage wasnt doing much good. I have bios 17.

EDIT: Ambient was 25C at the time of my tests. Should I be using prime to watch for which cores are failing or keep using OCCT?
 
Trial and error. Try some of the recommended bios settings I posted in red. Only way to find out. I suspect your MCH needs at least 1.3-44 range. Just noticed your PWM temps that sounds too high (with bolt mod) as well but a minor concern really. Just use OCCT since its fast and efficient. Need more info as well like what settings you are using. Memory 1:1? etc
 
I ended up with 3.4 8x425 stable and my memory is at 1063mhz. I could get 3.6 stable but I was putting 1.54V to get it there and my temps were right around 64C with PWM hitting near 80c. It just wasn't worth it. 8x425 gives me a higher FSB then 9x400 so after running benches comparing the two there is very little difference.
 
^ four hours of prime 95?

I've always heard 8 hours was the benchmark stability test! (overnight)
 
I will shoot for 8 hours tonight... My file server ran 14 hours stable at the same speed so I am not to worried about it not making it. I believe its 4 hours if you want to put your self into the Hard Forum Intel overclocking Database. Only thing I notice different is sleep function of Vista isnt working correctly. Might need more MCH voltage? System powers back up with no video. My dual core overclocked was fine. I am running the same MCH/NB voltage I ran with the dual core.. should I need to supply more for the quad?
 
Heh.... i got lucky to get 3.0 with my q6600 :( 9x333, 1.3625 V My setup does not like q6600 i use to have e6600 and had it 8*400 for last 2.5 years but no luck hitting same settings with q6600 i got yesterday :( Seems either my PS craps out or PWM can't handle Quad.

Bah... not to mention i'm all on Water.
 
Vlad_13 said:
Seems either my PS craps out or PWM can't handle Quad.
Click to expand...

Shit, no wonder. You have a Smartpower 2.0 PSU. Those suckers have a tendency to die under heat, light or heavy load, time or a combination of the three. I'm surprised it hasn't died under the strain of your rig.
 
Archaea said:
^ four hours of prime 95?

I've always heard 8 hours was the benchmark stability test! (overnight)
Click to expand...

Very true I've seen stability freaks going for 20 hours minimum. I've found OCCT to be equivalent to 4-5 hours of Prime95. 3-4 hours is tops for me. You can do the overnight thing but at the end of the day you just get what you already knew after 4 hours and a nice extra electric bill cost lol.
 
^ In my case my system is often stable for a four hour interval with my Q6600 and Abit IP35e in XP @ 3.6Ghz....but if I try to leave it running all night it's always crashed by morning because in the additional timeframe (past the four hour mark) it dies! It typically broke down at the 5-7 hour mark for some reason.

By my understanding an overclock isn't stable period if it EVER crashes in Prime95 or any other program. Overnight was typically a standard timeframe through the years (I thought) to pretty much say okay --- it made it a full night -- it should be able to continue on indefinately!
 
Archaea said:
^ In my case my system is often stable for a four hour interval with my Q6600 and Abit IP35e in XP @ 3.6Ghz....but if I try to leave it running all night it's always crashed by morning because in the additional timeframe (past the four hour mark) it dies! It typically broke down at the 5-7 hour mark for some reason.

By my understanding an overclock isn't stable period if it EVER crashes in Prime95 or any other program. Overnight was typically a standard timeframe through the years (I thought) to pretty much say okay --- it made it a full night -- it should be able to continue on indefinately!
Click to expand...

Other than Prime95 and the other handful of benchmarks what would tax your system for every single second for 4 hours? exactly. Comes down to what people perceive as stable. You got the extreme and the opposite.
 
I got 8 hours of Prime stable and didnt change a single setting. Last time I had just stopped it at the 4 hour mark. I let it go for 8 hours this time and stopped it again... I think im stable enough. Im not seeing any issues in games or applications.
 
Encoding four x.264 720p blu-ray rips will keep your cpu at 100% for a good 12 hours or so. That's been my stability assurance for my overclock.
 
i'm on 450 x 8 with 1.4125v on my ip35-pro and all is fine. i got a zerotherm nirvana to cool it all down.
 
Danny Bui said:
Shit, no wonder. You have a Smartpower 2.0 PSU. Those suckers have a tendency to die under heat, light or heavy load, time or a combination of the three. I'm surprised it hasn't died under the strain of your rig.
Click to expand...
Off topic, I had this PSU and it died on me once, then I rma'd it. Received a replacement and it died on me again a few months later.
On topic, do the vdroop mod by jonnybravo. My vcore in bios 1.45v, vcore idle in windows 1.45v, vcore load 1.45v.
 
Pharoahe808 said:
Off topic, I had this PSU and it died on me once, then I rma'd it. Received a replacement and it died on me again a few months later.
Click to expand...

Rock solid, 3 years on x1800xt and e6600 @ 3.2 with current setup (diff ram and +2 HDDs)

*Knocks on wood*

But i'm sure my problem is the PSU :( Under a heavy OC it craps out... meh.... I'm good at 9x333
 
I have a 40mm fan mounted on the PWM and a 80mm fan blowing air across the nbridge. I'm using a Tuniq Tower with an extra fan mounted to the back. Prime stable 24hrs and game stable completely on vista 64. Running bios 17. Q6600 is a g0

Edit: my gref is at 73%
2h4c2m8.jpg
 
Pharoahe808 said:
Off topic, I had this PSU and it died on me once, then I rma'd it. Received a replacement and it died on me again a few months later.
On topic, do the vdroop mod by jonnybravo. My vcore in bios 1.45v, vcore idle in windows 1.45v, vcore load 1.45v.
Click to expand...

How hard was the vdroop mod... I know that would get me stable at a lower voltage.. I watched as the temps on the PWM got higher the vdroop seemed worse. 1.54V I am pretty sure would get me stable for 8 hours at 3.6ghz but at that point its not worth the extra heat and noise to cool it. I have not tried 8x450 yet but I have a hard time believing I would get 8x450 easier then 9x400. Or am I wrong?
 
unrealsword said:
I have a 40mm fan mounted on the PWM and a 80mm fan blowing air across the nbridge. I'm using a Tuniq Tower with an extra fan mounted to the back. Prime stable 24hrs and game stable completely on vista 64. Running bios 17. Q6600 is a g0

Edit: my gref is at 73%
2h4c2m8.jpg
Click to expand...

Did the Gref make a difference? I have tried 63,67, 69 and 70 but nothing else.. so far 30 min stable is my best.
 
Tried your settings.. core 3 failed in 3 min. Took vcore up to 1.50V in bios..1.47 in windows. Ive never taken it over 1.53V in bios 1.50 in windows. How high is safe for a Q6600?
 
I was under the impression that 1.5(win) was the highest safe voltage for the Q6600/Q6700. I need that much to get stable at 3.6ghz w/ my Q6700 but I've gone to 3.8 for a few hours at higher vcores (out of curiousity). I'm on water though, so my temps never cross 52C load. My PWM temps hit 93C according to Everest though...putting a 40mm fan didn't put a notch in that...so right now I have a 120mm fan blowing at it, plus I did the bolt mod for the mobo heatpipe setup.
 
prasvt said:
I was under the impression that 1.5(win) was the highest safe voltage for the Q6600/Q6700. I need that much to get stable at 3.6ghz w/ my Q6700 but I've gone to 3.8 for a few hours at higher vcores (out of curiousity). I'm on water though, so my temps never cross 52C load. My PWM temps hit 93C according to Everest though...putting a 40mm fan didn't put a notch in that...so right now I have a 120mm fan blowing at it, plus I did the bolt mod for the mobo heatpipe setup.
Click to expand...

Yup did the bolt mod.. ive seen my PWM hit 80c under higher voltages but doesn't get much over 73C at 3.4ghz. Think I will just live 3.4ghz. 1.54V and I saw temps of 67C but it is 78F in my house so that doesn't help much.
 
I screwed in screws (6 inch long) into the pwm and added the fan on top of them (so basicly the screws are sticking in front of the exhaust fan and taking pwm heat out). Sorry that it wasn't stable for you :(
 
Copyright said:
How hard was the vdroop mod... I know that would get me stable at a lower voltage.. I watched as the temps on the PWM got higher the vdroop seemed worse. 1.54V I am pretty sure would get me stable for 8 hours at 3.6ghz but at that point its not worth the extra heat and noise to cool it. I have not tried 8x450 yet but I have a hard time believing I would get 8x450 easier then 9x400. Or am I wrong?
Click to expand...
I followed the advice in post #4 here:
http://www.ocforums.com/showthread.php?p=5317306
"I never needed a fine tip iron. Tin the tip of your wire, leave a little extra solder on the end. Touch the wire to the contact point, then heat it just above the solder by contacting the wire with your massive ugly iron. Solders right on, no sweat."

Extremely easy, but you have to be patient. I couldn't get it to stick the first try. But then again, YMMV. And don't apply too much heat for extended periods of time, you'll probably damage something. Try at your own risk!
 
You must log in or register to reply here.
Back
Top