1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Overclocking 3770K at stock voltage

Discussion in 'Overclocking & Cooling' started by Octopuss, Jul 30, 2012.

  1. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    could be board power delivery too..one core is far less load. Actually, that might also explain the high temps, if VRM on board was whacky...give the board a good visual inspection.
     
  2. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Well, that sounds logical too. However when I checked the errors in the log, it was always core 2 and 3 (at least I hope I read it correctly).

    When I enabled two cores, the errors were back. That's why I primarily suspect the CPU.
    I also disabled HT btw.

    Ill check the board anyway, of course.
     
  3. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Might wanna try HT on, too. There's really no need to disable anything on IVB for 24/7 clocking, and I have yet to see a game that has issues with HT in the past year or so, while for me, the temp difference is negligible.
     
  4. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    That was just a test you know. The PC is in pretty much useless state - I was just doing one last attempt at finding what was behind it. I will RMA it tomorrow anyway.

    edit: HT on, ramped up some frequency... and it's still stable
     
    Last edited: Aug 9, 2012
  5. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Do you mean that once you enabled HT, and all cores, it's fine, but HT off, it is not?
     
  6. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    No. I enabled HT and kept only one core enabled, and it was stable anyway.
    I have it currently running with 1.230V vcore.

    I am 90% sure it's the cpu, because I was getting BSODs and other errors in idle too.
     
  7. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Everything is RMA'ed and I will hopefully get a replacement next week. I only hope they won't tell me it works fine for them.

    P.S. I read some reviews about the cooler I wanted to buy and seems like I will have to start looking again. Weird mounting mechanism, crappy fan clip, and as a bonus I think I couldn't fit my memory under it, the modules are bloody 6cm high.
     
  8. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    lol I just got a call from a tech regarding the RMA. Seems like the board freaked out and fried one of the cores since a corner of the cpu was somewhat burned from underneath (I didn't notice that at all). Also he couldn't even get the board to work at all. Weird stuff going on.

    I will be getting brand new board+cpu tomorrow. Yay!
     
  9. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Ha, sad to hear, but at least it explains the problems!

    Let us know when you get the new hardware and we can try again! lol.;)
     
  10. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    lol yeah. He asked me "have you OCed the thing, just between us?" :D Sure I told him I did, but we agreed I couldn't fry it the way I was going at it. He said it looked like the board power regulator or whatever it was gone mad and pumped some insane current right into the poor cpu. Shit happens I guess, but it's like the first time I've seen something like this. Especially with nowadays' hardware.

    I really want to get a different cooler prior to assembling the thing again though. Tough choices! (still crying a river over the Scythe, totally unsure whether to try it or not).
     
  11. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Pro tip:

    Next time do not adjust power settings in BIOS for the VRM. Auto is enough to go way over 5000 MHz.

    ;)


    I think the board was knackered from the get-go though, as that would very easily explain the horrible temps you were getting since day one.
     
  12. Crap Daddy

    Crap Daddy

    Joined:
    Oct 29, 2010
    Messages:
    2,785 (1.71/day)
    Thanks Received:
    1,081
    From the start something felt fishy with the motherboard rather than the CPU since with auto settings it wasn't able to keep stable 4GHz. Anyway good to hear you'll get a replacement so soon.
     
  13. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Nono, the instability wasn't real, it was the damn bugged old version of Prime95.

    The temperatures were horrible. I saw people with the very same cooler hitting lower temps with 0.1V more, lmao, and was like WTF are you kidding me... Guess the board really was dead before I turned it on :p

    So that means I might not need a new cooler at all!!
     
  14. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Yeah, you should not. It's not going to be the best, but it's butt-loads better than the stock cooler, for sure. I'm using an H100, and still get pretty high temps at times, so some chips just run hot, period, but they should NOT be like yours was.

    That's why I kept offering more stuff to change, etc...it just didn't seem right.
     
  15. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    What did you mean by Next time do not adjust power settings in BIOS for the VRM btw? I don't have manual atm and still didn't memorize all the settings yet. Which one is this?
     
  16. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    This page, where you adjusted a few things. Really, you should not need to touch anything on that page.:

    [​IMG]
     
  17. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Ah! Well I followed the overclocking guide :p
    At first I only changed voltage response and vcore LLC though, only randomly flipped other around later when I was desperate.
     
  18. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    Wasn't a guide I wrote....:p

    Seriously though, you do not need to adjust anything there, and doing so will only really give you more heat to deal with. Xzibit mentioned this as well several posts back.
     
  19. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Everything is up and running.
    Default settings still, everything on auto.
    Testing CPU with Prime95. Temperatures are at 80°, and CPU-Z reports vcore to be 1.080V. rofl
    Quite a difference from 1.2+ on original board.
    How could that pass factory testing is beyond me...
     
  20. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta

    Your "extreme" setting for loadline explains the higher voltage. Exteme makes it so the CPU voltage increases under load, rather than going lower as it should. I wouldn't really call it that out of place, honestly.
     
  21. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    You're right, I just tested it and it really pumps more volts then. It is still lower than before with same settings!
     
  22. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    14,213 (4.31/day)
    Thanks Received:
    7,494
    Location:
    Edmonton, Alberta
    New CPU though, right? There is no set voltage for these chips, each is individual.
     
  23. Xzibit

    Xzibit

    Joined:
    Apr 30, 2012
    Messages:
    1,340 (1.24/day)
    Thanks Received:
    423
    There just a range of 0.25000-1.45000 i beleive

    For current phases on the motherboard:
    I would test with every settting on (Lowest setting avaible) "Regular" <- thats what my board says.

    J.J. from Asus (I know your running Gigabyte) said on a overclock demo. To start with that because the chip will only ask for the current window Intel has specified it for and the additional settings are for extreme overclocking where the chip is reaching its limits and your pushing it beyond the recomended envolope and you'd just be adding heat to the system.
     
    Last edited: Aug 11, 2012
    Octopuss says thanks.
  24. Octopuss

    Octopuss

    Joined:
    Aug 13, 2009
    Messages:
    1,193 (0.57/day)
    Thanks Received:
    193
    Location:
    Czech republic
    Hm, I am at 4.2GHz again, CPU-Z says vcore is 1.032V, and temps while Priming are... 78-85°. Shall I be afraid again? lol

    Didn't touch the PWM stuff this time.
     
  25. Xzibit

    Xzibit

    Joined:
    Apr 30, 2012
    Messages:
    1,340 (1.24/day)
    Thanks Received:
    423
    I'll assume your running the -/+ Offset and not the Auto voltage ?

    Once you figure out the lowest voltage to run a certain multiplier. It will then be a matter of how cool you can keep it with the cpu cooler.

    Voltage looks good @ 4.2ghz i was peeking 66c but I'm on a ALC and Cadaveca is on a better one also so we'll have lower temps.



    Just looked at your case and it might be a ventalation issue. If your area around the PC is just warm and the case Lian-Li PC-8NWX just has 2 120mm fan and its not pushing enough fresh air to the CPU cooler to circulate the hot air out thru the rear 120mm fan then it might be building up a little hot air pocket everytime you run prime and could give you that extra heat.
    If you installed the optional 140mm fan on top it might very well just be your not getting rid of the heat as fast with the cpu cooler.

    NOTICED: Its a top mounted PSU design.. Dont know how you have your Enermax MODU82+ facing but if you plan on leaving your CPU overclocked you might want have the fan facing up so it pulls in fresh cool air from the 140mm opening on top of the case. You dont want +80c temps being blown into the PSU if you have it facing down into the CPU cooler.

    Try leaving the case side panel off. That lowers my temp about 2-4c. Might be better for you since your cooling with air.
     
    Last edited: Aug 12, 2012

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)

Share This Page