Overclocking Q6600 G0 on Gigabyte GA-P45-UD3P motherboard

Well I have it up and running with a 333Mhz FSB at normal voltages and it has been stable with both Prime95+3dmark and Prime95+Linx running for a few hours (I keep switching back and forth). The only thing that I don't like is that Core's 1& 2 keep having their temps jump 7 degrees above cores 3&4 then drop back down.
 
When I was getting 'good' tests, my temps where within 3 C of each other and 2 cores each would have the same temps. Once they got their hottest, they didn't really vary a lot. When you're satisfied with your tests, you should run a few games and see if they fail. Mine seem to fail in 10-15 minutes.
 
My temps seem to jump all over the place from 62-71 for the first two cores and 62-67 or so for the other two cores. I will hopefully get a chance tomorrow night to play some games and beat on it some more.

My temps seem to be up about 10 degrees from a 266Mhz FSB although at idle it is running at 2Ghz up from the 1.6 it had been.
 
Is it possible that the extra 10C I am seeing while idle is a result of switching my video card from a 8800GTS to a 260GTX the day before working on this overclock? I hadn't tested the temp after switching cards.
 
Is it possible that the extra 10C I am seeing while idle is a result of switching my video card from a 8800GTS to a 260GTX the day before working on this overclock? I hadn't tested the temp after switching cards.

I doubt it. While it is strange, I wouldn't worry about temps like that unless you're having stability problems.
 
I wouldn't worry about the difference in temps between cores so much as the overall temps in general. Those temps are kind of high for only running 3Ghz. What cooler are you using again? Do you have your vcore set manually or is it on "auto"?

Maybe reseat your cooler. Check case airflow. Something is going on that you are seeing 70c+ at 3Ghz.
 
I am using an Arctic Cooling Freezer 7 Pro. My Vcore is set for "Normal" which is 1.3volts although while watching CPU-z I kept seeing this drop down to 1.24 or so volts. The airflow may be a bit inadequate because I have not figured out a good place to stuff the extra cables from the PSU and the long video card is probably blocking a bit of the air flow from the intake fan. The CPU was only hitting 71 for a few seconds before dropping back into the mid-60's.

What has me worried is that it is only the first two cores which get that high BUT that could also be because all of my OS functionality is on those cores also.
Is 70c from Realtemp realistically that bad since it is still showing around 30c before critical?
 
I doubt it. While it is strange, I wouldn't worry about temps like that unless you're having stability problems.

The 3Ghz overclock has only been running since last night but it seems stable. I forgot to run a Prime95 test overnight (I don't want to run it with LINX unless someone is around to monitor it every once in awhile).
 
I might undo the OC and see what my temps are reading. That way I will know whether the higher temps are due to the OC or the 260GTX.
 
Where can I specify the PCI bus speed when overclocking? I found the setting for the PCI-E speed.
 
Where can I specify the PCI bus speed when overclocking? I found the setting for the PCI-E speed.

Exactly...the PCI bus speed is the PCIE speed, it's just not labeled as such. It's in the first section near the top, underneath the multiplier and all that stuff. Set it to 100.

In my original quick n dirty OC, it was a 333mhz 9x multiplier for 3.0ghz and other settings on 'Normal', I was achieving 67 C in Prime95. There's articles here that say decreasing the multiplier and increasing the FSB should show a noticeable difference in temperatures. That's why I was going for a 400 FSB 8x multiplier to equal 3.2ghz. Not only was the CPU running faster but the temps were much less...as I've stated before, <30 C idle and mid 40's with Prime95.

You said your Vcore was 'Norma'l. Do you mean you actually set it to 'Normal' or you have it set to 1.3? Reason is, if it's on 'Normal', I wouldn't be surprised if you set it to 1.3 or less and see lower temps.

IMO you're in a better boat than I if you're games aren't crashing :(
 
I had it set to "Normal" not a specified voltage. I haven't tested a game yet but did run 3dmark Vantage while running Prime95 (after a few hours of running Prime95+LINX). My temps ranged between 62C and 71C on full load with only the first two processors going above 66C or so.

Do you still have the links about using 400Mhz FSB to lower temps?
 
Maybe I will give 355 another try tomorrow night and raise the MCH +.1 to see if the speedburn will run successfully without causing an error on reboot.

Its the oddest thing since I had 3 different burnin tests running from the UBCD without errors but when I rebooted after running one of them my MB would whine about problems with the overclock.
 
Hey Ryland, let us know your results - I just picked up a GA-P45-UD3P to run my Q6700 on and I've been trying to get it stable @ 3.5 (Tuniq Tower 120 w/ extra push fan) for folding@home smp usage. I thought I had it working @ 1.472V yesterday but after 24hrs it froze. So this morning I upped the Vcore 1.488 and lowered the MCH to 1.30 (it was high and I'm just running 350 x 10). Anyway I turned on one smp instance of folding@home and left for the office - so hopefully it's still running.

Bill - if you're still looking at this thread, could you give some advice on CPU Vtt settings for 65nm Quads? I know not to go past 1.4V on my Q9650 but I'm not sure what I should set the CPU Vtt at (if at all) for the Q6700.

Aside from that, I set the CPU Reference to roughly .667 of CPU Vtt (which should help in general from what I've read on GTL ref voltages).
 
Is it possible that the extra 10C I am seeing while idle is a result of switching my video card from a 8800GTS to a 260GTX the day before working on this overclock? I hadn't tested the temp after switching cards.

Yes, the back plate of the 260 gives of extra heat, same happened to me when i installed 280 in my system.

My q is at 3.0GHz 333MHz FSB. 1.3675v

For the stability test, try Intel Burn Test with FurMark Xtreme Burning Mode. This puts an incredible stress on to the system.

Good luck
 
I had it set to "Normal" not a specified voltage. I haven't tested a game yet but did run 3dmark Vantage while running Prime95 (after a few hours of running Prime95+LINX). My temps ranged between 62C and 71C on full load with only the first two processors going above 66C or so.

Do you still have the links about using 400Mhz FSB to lower temps?

I don't know which thread it was in. You can try to search around but if I find it I'll let you know. It might of been in the "How to OC" sticky.

Ima try a quick n dirty OC 9x 355 FSB and report back. 3x mem speed like you said. All other settings normal. If you have the same settings and aren't crashing (yet), then maybe this will help prove if I'm just retarded or if there's something wrong with m setup.
 
Do you still have the links about using 400Mhz FSB to lower temps?

I found some of the info HERE

You'll have to read through the BS, but here's mainly what I was looking for:

o illustrate, consider this analysis of two difference vcore settings and the temps they produce on my Q6600 @ stock settings (9x266=2.40 GHz) as well as running overclocked (9x333=3.00 GHz). The two voltages I used were 1.1375V and 1.2625V set in the BIOS that correspond to the two clock levels of 2.40 GHz and 3.00 GHz respectively. In case you’re wondering, these translated into 1.112V and 1.232V in Windows as read by CPUZ.

Prime95 ran for 30 minutes and the temperatures were averaged over the last 10 minutes of those runs (well after they stabilized). Room temp was 75-76 °F. Notice that the difference in voltage is ONLY 0.120 V or 120 mV, but this seemingly small difference brought the load temps up by an average of 6-7 °C per core!

Code:

Run1 (9x266 @ 1.112 V), Average temps (°C): 51,52,50,50
Run2 (9x266 @ 1.232 V), Average temps (°C): 57,58,57,57
Differences (°C): +6, +6, +7, +7

Now if I add a faster FSB, they increased further:

Code:

Run3 (9x333 @ 1.232 V), Average temps (°C): 61,61,60,60
Differences from lowest voltage (°C): +10, +9, +10, +10
Differences from same voltage (°C): +4, +3, +3, +3

I forget where exactly I read 'lower multiplier can mean lower temps', but the info is out there. Shit, try it for yourself. Write down your temps @ 9x with current FSB, then set the multi to 8x and up the FSB to equal same CPU speed. Even with higher volts, you should see a drop in CPU idle temps as well as max temps.

Of course, don't forget what you're current 'stable' settings are.
 
Ryland, you smart son of a bitch! No promises yet...BUT...you may be on to something here!

I set a 9x @ 333 FSB = 3.0ghz...
PCIE 100....
mem at 1066 (max) w/ 3.4 mem multiplier and mem voltage of 1.31 (recommended for Dominator 5 5 5 15 1066mhz)
Vcore 1.38ish
All other voltages set to 'normal'

and you know what I got?

No crash for 45 minutes of hardcore Crysis AND and...oh I don't know, another 10 FPS...on top of the 20 I said I had before vs default 2.4ghz

So you know what I think? 9x@333FSB FTW!

BTW.....the Gauss cannon in Crysis:Warhead is fucking WIN.

Can I top that? Well, how about 28-32 C idle muthafucka!

As I said, no promises yet....but I'll hit up some L4D now. That always locked up guaranteed. I don't know...I was running no more than 1.35 Vcore with the 400FSB...maybe it wasn't enough? I mean, I'm on 1.38ish right now with only the 333FSB. Plus I didn't adjust ANY other voltages, except for mem. Maybe 1.35 wasn't enough for 400FSB.

L4D results soon...
 
OK, so I don't know what it is with L4D, but the game seems glitchy to me. Even with no OC, I've had it get goofy, even with turning down my GPU clocks, so L4D probably isn't the best game to be testing.

Anyways though, within a few minutes the game got stuck, I minimized, and then it was fine. I turned shit down and it did it again. I turned it down again and then it was fine from there. Something about those settings that my computer just doesn't like. I know the QXSAA whatever the hell that AA is called, freezes my computer in no time. I've got it on bilinear and 4AA and it ran fine I guess. Soo I'll have to find another game to test later.

But still, Crysis didn't freeze so that's a major plus.
 
I put my system back to stock speed last night so that I could check its temps and found that it is still running about 8 degrees hotter than when I had the 8800. I need to figure out how to increase airflow in the case (plus I need to grind some of my drive cage off because it is resting against the 260). I may be able to move one of my harddisks out of the way to open up greater amounts of air towards the CPU.
 
I saw no significant difference in temps between using an 8 or 9 multiplier. Whether is was 8x400 or 9x355. Though I heard that held true more with nvidia chipset based boards. All I know is my P35 board doesn't do squat in terms of temps when switching multipliers.

No, actually, the NB ran slightly warmer with the higher FSB speed.
 
No, actually, the NB ran slightly warmer with the higher FSB speed.

I would expect it to actually. It seemed strange that a higher FSB speed would run cooler with a lower multiplier (of course I don't understand where the multiplier works into the equation...)
 
I would say it's going to depend on the motherboard whether you'd see a difference or not. I would also expect the NB to run slightly warmer with a higher FSB. I would expect the CPU temps to be cooler though. Maybe you'd see a drop in temps at a 333 FSB. With the Antec 900 I have, the HSF is blowing towards the back right into an exhaust fan, which moves a higher volume than the HSF, so I would expect my rear exhaust fan is pulling some of heat off the HSF, which helps temps a lot.

Ryland what kind of a case do you have?
 
I have an Antec Solutions. The Artic Cooling fan is blowing towards my exhaust fan on my case but the intake fan is in a bad spot (cools the HDD's but the video card blocks too much of the airflow).
 
Hi all. Good thread for me. I am attempting to get my Q6600 to 3.20 on a GA-P45-UD3P.

I wanted to again ask a question that BigBeef posed a couple pages back but I don't think was answered.

There is a setting in M.I.T. called "Load-Line Calibration"

********* Motherboard Voltage Control *********
>>>>>> CPU
Load-Line Calibration.................: [Disabled] - Default setting in "Optimized Default"

[Disabled] is noted as "Follow Intel spec"
[Enabled] is noted as "Improve VDroop directly"

Should I leave this at default disabled or enable it when overclocking to 3.0-3.2?

Thanks!
 
A quick google search brings up that that reduces VDroop so keep it enabled.
 
I ran some stress testing both with and without overclocking. At idle both were about the same temp and at full load the overclock was only 6C hotter than normal, maxing out at 69C on 2 cores and 65 on the other two.
 
Ugh...my antivirus is giving errors when emails get opened so now I don't know whether its from the OC (which has passed every test I have thrown at it, albeit I haven't left it testing overnight yet). I have my wife running a 15 iteration of IBT now but it passed that twice yesterday.

Edit: It passed 15 runs of IBT but its possible that something happened yesterday to corrupt Avast.
 
I've used AVG for years with no issues.

I played some NecroVisioN last night. SWEET game made by 1C, the group that also made Cryostasis. Runs off DX10 so you know shit looks bananas. Played for about an hour with 60-100 FPS cranked, no lag, no skipping, and no crashing whatsoever. I think tonight I'll try to lower some voltages.
 
I currently have my voltages set to Normal (by selecting the actual voltage that matches, not selecting "Nnormal"). I also disabled Load-Line Calibration which resulted in cooler temps. The weirdness with Avast could just be something unrelated to the OC or not....I have no idea at this point.
 
Ahh screw it..I think Im gonna shoot for 3.2Ghz tonight pushing the voltages up slightly as Bill said to try.
 
Thanks guys.

I've enabled "Load Line Calibration".

@Ryland, I too am thinking of following Bill's advice and hitting that 3.2 "sweet spot."

I will post results after some adjustments and testing.

Best of luck!
 
Well on reboot it seems that Avast is now behaving itself. Im wondering if my wife might have hit some bizarre key combination on the 2nd keyboard (PS2) which caused Avast to go crazy...
 
When in doubt, blame it on your wife.

Worst case scenario would be that you can't find a stable 3.2ghz OC but you can find a stable 3.0ghz OC, which is very acceptable.

When I was using the 8x multi with 400FSB, at one point I had set all settings to specified normal voltage and that didn't work either. But like I said a page or two ago, my Vcore may just not have been high enough. I mean, CPU-Z says 1.344v with my 9x@333FSB, and before I was trying to run the 8x@400FSB with 1.35v...at some point I'd like to try the 3.2ghz again.

So if you're going to push your CPU higher, make sure you jack up the VCore :)
 
I think I will hold off on shooting for 3.2 until after I fix the airflow issues that the length of the 260 is causing since I am already hitting 69C on 2 of the cores. I still don't like the fact that I have 2 cores hitting a max of 4-6C than the other 2 cores. Im not going to reseat my HSF so I guess as long as they don't go critical its not an issue.
 
I'm taking a look at your case on Newegg. IMO it doesn't seem to provide very good air flow. There's only 1 fan and it alone can't pull out all the warm air in your case. Plus there's nothing blowing cool air in.

You say 69C idle or during P95? 69C is getting quite hot but if it's only during Prime95 running your cores 100%, then I wouldn't worry about it. Prime95 tells you how hot your cores will get IF they were running maxed all the time, but how often do you really plan to do that in every-day computer usage? If you're getting 40's in-game, then I wouldn't worry about it. Right now I seem to be 28-32 idle, around 45C in 3D Mark, and 59-62 C in Prime95. The reason I don't care is because not even Crysis will make my CPU run that hot.

If you're getting 69C idle, then I would say there's a problem. Either your case is too restrictive or you need to reseat your CPU. I'm compulsive when it comes to applying the thermal paste...it must be done right to keep temps as low as the paste possibly can.
 
I just ran the hour long OCCT test and it only hit 61C without errors.

It was 69C at full load with Prime95 and another test running at the same time.

The case I have has a 120mm intake fan and a 120mm output fan that the HSF blows into.
 
Yeah that's good. I wouldn't worry about temps them. I just played a round of L4D and was in the mid 30's C.
 
Well I played an hour of Crysis Warhead without a glitch...Not too shabby. It has my system as Gamer.
 
Back
Top