Q6600 @ 3.8 GHz air cooled

Gigabyte GA-P35-DS4
Intel Core 2 Quad Q6600 (G0) @ 3.8 GHz
Arctic Cooling Freezer Pro 7
1.5v cpu voltage set in the bios
Idle temps 34/36/33/33

Surprise anyone?

45 answers Last reply
More about q6600 cooled
  1. show me a 12 hour orthos stress test and I'll be impressed. So far so good though....
  2. Very good idle temps for the AC7.

    It would be interesting to see load temps though, although I'm expecting 70C+.
  3. hehe, 1.472 in windows should bomb in small ftts judging by the VID alone, if that even matters! I need my VCore to not droop below 1.472, hehe, or it bombs at that speed! So its 1.5250 in the Bios now, and 1.50 in windows.

    Good luck with those tests!

  4. For real, I can pump 1.6v into my 6750 and hit 4.0 long enough to start CPU-z and take a screenshot :p

    Until we see that it is 24/7 stable at that speed, it's not impressive :p
  5. Given his CPU VID I imagine he can run 24/7 stable at 3.8 with that voltage or slightly higher.
  6. I'm sure he could, with adequate cooling. The AC7 isn't a bad choice at all, but I'm not sure it will handle the load temperatures you'll be pushing at that frequency. Also, he didn't mention if he calibrated his thermal monitoring yet, so we have no idea if we can trust those idle temps.
  7. Mine is a 1.2625! And honestly I think I am getting some other sort of error when I go over 3.8. It always Blue screens at the same spot in prime95, and thats when it switches from the first set of tests three to four.

    Once it goes to complete test 3 and begin four, thats where it always bombs. I havent really toyed with it yet, because its fine at 3.8, but still, what could it be? It fails from its stable vcore, on up like 7 notches. 1.5250 up to like 1.5750. It has no effect. Fails all the time, havent had a blue screen yet from normal use, even playing crysis and such.

    Maybe my other voltages need to be adjusted. Maybe my processor is at its OC limit.

  8. The higher you try to push the frequency the more voltage is needed per jump. While you may be fine at 3.8 with 1.4 vcore, it could take 1.7-1.8 to hit 3.9. That's just the way it works. 3.8 seems to be the very max you can get away with on air, but it's going to be very warm. If Cnu won't keep his at 3.8 24/7 with such a good sample, that has been lapped (and the HSF) then you know it might be a little too much.
  9. That holds true! The more speed, the more power. Is anyone else interested in the VID, and if the lower ones pretty much automatically get to oc better than the higher vid ones, and stay cooler and at less voltage while doing it?

    Maybe someone should start a pool, or some such here and people post there coolers, speed and voltages, and their factory VID with all the junk disabled that change speed and such, that could fool core temp.

    So far just from what I have been reading, it does look like heat, thanks to more voltage off the top, does play a factor in OCing, and that damn VID seems to keep your heat down when its down!

    Got me.
  10. what does VID stand for? I'm guessing it has something to do with voltage, idle, and drop?

    Anyway, I OC'ed my q6600 (go) w/ the AFP7 to 3.3 ghrz@1.325v. Ran Small FFT's for 2.5hrs before an error. I've got a 8800GTS OC'ed to 800mhz and it's the bottle neck in Crysis. I tracked my core temps while playing Cryses maxed, and they're 10C cooler than when running FFT's.

    So you can OC to 3.8 and it might play Crysis w/o crashing but I doubt it actually performs any better than 3 ghz b/c the processor is not the bottleneck (which is why it doesn't crash in the first place).
  11. Wow if that is stable that is a great setup. 3.8ghz Quad? The damn Q6700 is $530 Retail. Its amazing how much money you can save and the performance you can get with a little knowledge. 3.6ghz quad on air is sexy, let alone 3.8ghz IMO.

    Does anyone else think those idle temps are a little low for 1.5v?
  12. "Also, he didn't mention if he calibrated his thermal monitoring yet, so we have no idea if we can trust those idle temps."

    maybe I'm missing something...how am I supposed to "calibrate my thermal monitoring"?
  13. Yes, that's what I'm trying to get across to him. I highly doubt he calibrated his thermal monitoring properly.
  14. Oh, I dunno, read the sticky in the CPU section titled "Core 2 Duo and Quad temperature Guide".

    You absolutely cannot trust any thermal monitoring out of the box. While you may be very lucky and be somewhat close to getting accurate readings, some can be off as much as 15c.
  15. VID is just the Voltage ID of your particular processor! It has a Factory rated one that prolly means it passed their rigorous testing at a lower voltage for whatever reason associated with the randomness of millions and millions of tiny little transistors.

    Guess that goes with all the Errata that are in a processor build. Some just do things better for whatever reason, so some pass their testing at a lower voltage, and get the VID stamp to prove it!

    Then again, maybe not. Intel is tight lipped about the VIDs. I think i will call up and harass them again! Basically its pay pack!

  16. Thanatos421 said:
    Oh, I dunno, read the sticky in the CPU section titled "Core 2 Duo and Quad temperature Guide".

    You absolutely cannot trust any thermal monitoring out of the box. While you may be very lucky and be somewhat close to getting accurate readings, some can be off as much as 15c.

    That is SPEEDFAN that needed the 15 degree adjustment. Both everest and coretemp are know to be accurate within 1c.
  17. Not always perp. You still need to go through the calibration process to be sure. My Core Temp was about 10c off, while Everest was pretty close. But not everyone wants to pay for Everest right?

    It is ALWAYS a good idea to calibrate even if you think it's accurate. Why NOT do it is the better question?

    If I pay upwards over $1000 for a new computer (or whatever you happened to pay) I'm not going to trust some freeware that some guy threw together in his basement. I like to know 100% what is going on in my system, and maybe that's why I've never had one fail me.

    The OP is on crack if he thinks that with 1.5v@3.8 and only the AC7 he will keep that quad for long, he's sadly mistaken.

    Why not save us, and him the greif and have him do this the RIGHT WAY and be 100% sure he's safe.
  18. I finally got my Maximus Formula back from Asus! Hey, wait a sec. Is OP original poster or other poster?? I am at 3.8 @ 1.5 volts as well! Waaaahhhhh! Wait, but I have a TRUE. Impressive thing that it is.

    Any pointers for the maximus formula Thanatos? How is it in relation to the p5k-e?

  19. Perp said:
    Given his CPU VID I imagine he can run 24/7 stable at 3.8 with that voltage or slightly higher.

  20. I hope you guys let us know when your chip blows. I've been dying to try to make a chip pop. :whistle: Someday!
  21. @Lupi: I can't comment on the P35 chipset, but others seem to love it. I have had 0 problems with the X38 Maximus SE though. It's a dream board. You can probably get good arguements for both chipsets. I guess in the end, it's personal preference.
  22. Dunkel, this is nothing! My errors alone top my running state. Booted into vista at 1.8125 VCore on my b3. It was hot, but operational! And I didnt have my water block on properly due to the capacitors being like a zillionth of an inch higher than the IHS. Temps when checked in the bios for the 10-15 seconds it stays on were 100-106 c.

    Soooo, if that doesnt make it simply melt down, dont know what will. Thats why I am trying to find out!!

    Thanatos, I finally got mine back, thats why I was asking. I will toss my B3 in it tonight so I can get a feel for hos its gonna run!

  23. lschmidt said:
    "Also, he didn't mention if he calibrated his thermal monitoring yet, so we have no idea if we can trust those idle temps."

    maybe I'm missing something...how am I supposed to "calibrate my thermal monitoring"?

    yea.. i think I am asking the same question..

    What exactly do you need to do.. aka.. steps you need to do to Calibrate ones thermal monitoring?
  24. Read the sticky in the CPU section of the Overclocking forums. It's titled "Core 2 Quad and Duo Temperature Guide", written by CompuTronix. It's a great guide and worth the time to read and follow.
  25. cnumartyr said:

    What vcore did you have to set it @3.89GHz?
  26. How high can the Q6600's temp go and still be safe?

    When I OC mine to 3.6Ghz...I break into the 70C+ temps w/small ftts in prime95.
  27. Evilonigiri said:
    What vcore did you have to set it @3.89GHz?

    1.576 Idle.

    To the other guy, read Comp's temp guide.
  28. cnumartyr said:

    Really? Are basing this off your extensive research on VIDs and corresponding overclocks?

    While I didn't compile the results myself I did read through about 100 pages on two different forums of people collecting VID + Overclock results for the q6600. 3.8ghz for his VID rating is not out of the question with his current vcore or a touch higher.

    On the other hand I don't think he can run 3.9 without really ramming up the voltage on his cpu which while it might work and with heafty cooling might run "just" cool enough, but it will start to die at an accelerated rate from the overvolting.
  29. Maybe I should explain a bit better.

    VID = Voltage ID = voltage your mobo will assign at stock speeds

    Intel lists the VID for the q6600 as .85 - 1.5 with 1.2 - 1.375 being the more common VID distribution with a few sub 1.2 VID making their way out of the intel factory.

    The actual impact is dramatic between a VID of 1.2 and 1.3.

    A chip with a VID of 1.2 can usually run stable at 3.6ghz@1.3v

    Compared to a chip with a VID of 1.3 that only runs at 2.4@1.3v

    That .1v difference in VID makes a huge difference in overclockability for both voltage required and in turn heat generation.

    I tested this personally on my q6600 VID 1.2375 which runs stable 3.6ghz@1.345.
  30. Thanks for the clearification Perp, I have no idea what VID is.

    Perhaps you'd like to define what you mean by stable as well. Our definitions may differ and your stable might be a joke to me.
  31. LOL, I love it, and yes, I believe the VID also makes a big difference in your OC. Its like being able to use extra voltage notches that the other doesnt have! I have a 1.2625 g0, thats a upper mid VID, I wish I had a 1.2 or lower one as well. I have a high VID b3. Its a flamer at 3.2! Its 1.3125.

    The VID is the voltage that it passed intels rigorous testing at. Some pass low for some reason or another, and thats how you get the VID. The highest VID they have released in the q6xxx series is 1.3250 And you DO not want one of those if you can help it! Its random, and no way to know what one you'll get since they all work at rated speeds and such for at least as long as their warranty period!

    Thats why the processor is rated at .85 - 1.5. They make real low ones that work for whatever reason at low voltage, but max out at 1.3250 to stay withing the 1.55 volt maximum rating for the q6600. Nice, huh? But Luck has alot to do with the one you'll get! I will be watching for a low vid one myself!

  32. cnumartyr said:
    Thanks for the clearification Perp, I have no idea what VID is.

    Perhaps you'd like to define what you mean by stable as well. Our definitions may differ and your stable might be a joke to me.

    Gaming stable - 1hr OCCT (2-3 times with no errors)

    24/7 stable - 12hrs prime95 small ffts (2/4 threads for duo/quad) and another 12hrs large ffts for memory testing

    Maybe not the most extreme stability testing, but I don't build servers.

    Side note:

    I like super-pi for a quick stability test, but I'm running 64bit vista on both of my machines. Super-pi doesn't function properly with these OS due to the aero effects and oddly vista sounds. Don't have the links to the issue, but supposedly disabling 2-3 things in the higher end vistas will keep superpi from crashing. I can assure you though that with both of my OS (64ultimate, 64homepremium) it crashes even at completely stock clocks. Sucks, because I really like the program as a general benchmark.
  33. Super pi runs fine on my vista 64!
  34. Odd, mine won't run on either machine with stock settings on either of my machines.

    I was curious about the issue and a few threads came up where people isloated the problem by disabling **** in vista. They probably aren't even sure what's causing it, but it doesn't seem to be isolated to specific mobo/chip combos. The only common factor identified is vista's with aero effects.
  35. It appears to run just fine over here without any mods to anything. Got me. As a matter of fact, it works on both my 64 bit comps.

    I agree that its always possible for something you use as a test to have problems like everything else does! Maybe that version needs you to disable driver signature verification?

  36. Did you have to increase the FSB voltage any? I'm impressed that you got it to even boot!
  37. Got AIM?
  38. Say Perp, what program are you using to figure out the VID? Coretemp?

    If so, I have come across people who's VID displayed in Coretemp changes from mobo to mobo, leading me to believe that what Coretemp displays is rubbish.

    It is also entirely possible that just by chance you happen to get a Q6600 that displays low and OCs very well. Just one sample won't prove anything. Furthermore, data on correlation between VID and OCs doesn't imply causation.
  39. Thats true! Just waiting to at least prove that it coincidentally works for me when I get another quad of any VID! (And if it works for me, I am mostly happy!)

    By reasonable deduction, one must conclude until seen otherwise that it would help with OCing. Simply because its is so low, making it cooler, prone to a longer life, and more than likely you will really notice a longer life at the higher speeds simply because you dont need as much voltage as the automatically higher VID ones would.

    But then again, it could just be coincidence. I would hope that the people that have shared their VIDs and temps and such would lie and say its cooler, not hotter, so when they have all turned out hotter and needing more voltage. ::Shrugs!:: But ya never know. I will have another quad within a month. And since I have a g0 and b3, doesnt matter the type to begin my tests, and since I have 5 motherboards, and my VIDs NEVER change... my data should be sound.

    ::Sticks tongue out at Evil.::

    So far its the normal reaper without those huge heat sinks, or balistix which hit 5-5-5-12 @ their rated voltage and 1066. Dunno if they could maintain their 4-4-4-12 at 950, but they look good!

  40. There are known to be a couple of people who've said coretemp is reporting vcore instead of VID but I don't know which hardware is having this issue. Pretty easy to tell if it's reporting vcore though, and the vast majority of people using coretemp are getting correct reading.

    VID is a value assigned by intel and it's what the CPU tells the mobo to set the "auto" vcore. The different vcore values at "auto" from mobo to mobo are rounding/voltage values from the mobo itself not the VID. I have never seen anyone say the coretemp reports a different VID on the same chip on 2 different motherboards.

    Also, lower VID values are definitely an indication of how far you can overclock a chip because of voltage headroom. There are lots of other factors in overclocking a given system (NB, RAM) not related to the CPU itself which can drastically change how far you are able to push the chip.

    There is far more than "one" sample correlating VID value to maximum overclock and the data overwhelmingly suggests that all else being equal lower VID = higher overclocking potential.
  41. I agree. I havent seen a single person saying Hey man, you all are a bunch of asses, because my 1.3250 VID processor does, blahblah.

    I am inclined to believe that out of all these people, though it really isnt ALOT of people, no one has reported good/excellent OCing with a high VID chip. As a matter of fact, all seem to be very hot and take a huge amount of vcore.

    I am right there asking every new guy that needs help for his VID so I can watch how their OCing proceeds. I will pay special attention to high and low VIDs. I wanna see someone with a 1.2000 or lower one OC.

    My guess is with that extra voltage you can use, OCing higher is a snap, unless its that unlucky Processor that simply doesnt OC. If a 1.2625 can hit 3.6 @ 1.38 volts loaded and stay very cold, under that 65. I think it was 62c.

    I can only imagine if I had access to that some odd 10-12 voltage steps due to the VID being low. I'll bet that a 1.2000 Vid could hit 3.6 at 1.32 volts in windows!! Thats my b3's VID, hehe.

  42. Of course I don't disagree with the fact that lower VID brings better OC, but I'd like to see more definite proof, perhaps a test.

    While data means a lot of things, it doesn't mean everything. Ex: The bigger the shoe size, the better the reading score. The correlation between these two variable is nearly 1, thus one would be inclined to think that bigger shoe size equates to better reading skills. However, there is a confounding variable, which is age. Obviously as you grow older, your shoe size increases and you grow smarter, thus the very close relationship with shoe sizes and reading skills.
  43. ::Scratches his head.::
Ask a new question

Read More

Gigabyte Overclocking