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

What is the average overclock on air with a 3770k 24/7 use

Discussion in 'Overclocking & Cooling' started by trt740, Jan 16, 2017.

  1. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    Mine does 4.4ghz at 1.26-1.27v. It does not seem to be the greatest chip ever, but I was wondering if this is an average overclock on air . Mine does very well on the lower end but needs about 1.35v to get truly stable at 4.5ghz and that seems a bit high from what I have been reading. I have only had this chip a few weeks. I have my cstates on and load line is on extreme and all the other setting are set right per the guides I have read. From what I read, most chips do 4.3 to 4.6 from 1.8v to 1.32v depending on the chip and cooling.

    The guide I used. Its a sandy guide but it still applies.

    https://hardforum.com/threads/official-asus-p8p67-series-overclocking-guide-and-
    information.1578110/


    Excerpt:


    4.7GHz or Below
    All bios values to Auto except for those noted-
    CPU Turbo Multiplier
    Dram Voltage to specification
    Internal testing has shown Auto Values will allow for stability in all forms of testing including high synthetic high load applications (Linx, Prime95, Occt ). In the event general stability is not achieved in these synthetic high load applications, you can adjust the noted values for improved stability.
    Digi + VRM options
    VRM frequency to 350 – Requires setting to manual adjustment and entering the specified value.
    4.7GHz and above
    CPU Turbo Multiplier – To desired value
    Dram Voltage to specification
    Digi + VRM options
    VRM Frequency change to 350 – this value will allow for scaling to 50+ multi without issues
    Phase Control change to extreme – this value will allow for scaling to 50+ multi without issues
    Duty Control change to extreme - this value will for scaling to 50+ multi without issues.
    No other values need to be changed. Unless otherwise noted all other UEFI values used are AUTO.
    K series overclocking and its affect on subsystem performance especially HD performance
    Quick note regarding options that can affect subsystem performance
    It is NOT advised to make adjustments to Cstates as this can considerably affect hard drive throughput performance ( especially SATA6G SSD or Sandforce 2 based SSD ). It is recommended that all CPU power configuration states be left on their default parameters. Overclocking tests have shown internally no increase in multiplier scaling when adjusting these values. * under special cases with high multi capable CPUs and synthetic high load applications ( Linx, Prime, Occt ) it may required C states to be disabled. This has generally only been confirmed for some 51-54 multi capable CPU’s.
    K series overclocking benefits from adjustments to Digi+ VRM options
    Advanced Digi+ VRM options recommendations
    VCCSA
    This has shown in internal testing not to improve overclocking, yet may still maintain the same level of stability while being lowered from its default value in order to keep the CPU cooler.
    VCCIO
    The adjustment of this voltage may help to slightly improve the overclocking capability of the IMC / DRAM, even though the default voltage is enough to run at a 2133MHz DRAM frequency. A 1.20v setting is more than adequate to maximize Memory overclocks in most cases.
    CPU PLL
    This setting did not improve overclocking, yet the user can maintain the same level of stability while lowering its default value in order to keep the CPU cooler.
    PCH Voltage
    This setting did not improve overclocking, yet the user can maintain the same level of stability while lowering its default value in order to keep the CPU/PCH cooler.
    PCH PLL
    This setting did not improve overclocking, yet the user can maintain the same level of stability while lowering its default value in order to keep the CPU/PCH cooler.
    K series overclocking and how BCLK is affected and how to best optimize BCLK scaling
    Quick Note for BCLK tuning
    BCLK clocking is considerably reduced with this new CPU architecture and as such it is recommended to focus on Turbo Multiplier adjustments for overclocking. With that noted should you choose to make adjustments to the BCLK, ASUS has enabled extensive control to maximizing scaling.
    Internal testing shows BCLK ranges from 102 to 109.1 with the largest level of scaling dependency placed on the quality of the CPU. CPU Frequency and speed is also directly tied to BCLK scaling. Internally we know that BCLK tuning largely depends on the CPU with approximately but we have noticed up to a 20% dependency on scaling potential based on the board design and PWM utilized.

    At 4.9GHz, generally the maximum clock speed will not change with a different combination of multi and bclk in this particular example.
    Example-
    100x49 = 4.9GHz
    103x48 = 4.9xxGHz
    Generally the CPU will not do much more than what your maximum CPU multiplier range is in most cases. Keeping this in mind, we offer very fine adjustments down to 0.1x increments to allow for the maximum BCLK tuning.
     
    Last edited: Jan 16, 2017
    10 Year Member at TPU
  2. Recon-UK

    Recon-UK

    Joined:
    Jun 21, 2016
    Messages:
    1,894 (3.87/day)
    Thanks Received:
    1,710
    A bit lower than Sandy on average but a good chip will do well.
     
    trt740 says thanks.
  3. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    That is what I thought. It still seems to be very quick. I was hoping for a bit a higher (45 to 4.7), but It might be my motherboard. It is a decent overclocking, but I would not call it a top P67 board. It is more of a mid level board built to last. Im still toying with a another attempt at a 170 or 270 board and a newer chip but for now it seems more than enough. (for no reason other than wanting something a bit newer). It is benching sightly above a 4790 at 4.0ghz. If you see anything that needs added to the setting other than lowering the PLL because I did that let me know. It did help a bit at 1.5v from 1.8v let me know.
     
    Last edited: Jan 16, 2017
    10 Year Member at TPU
  4. DOM

    DOM

    Joined:
    May 30, 2006
    Messages:
    7,626 (1.83/day)
    Thanks Received:
    863
    Location:
    TX, USA
    Does the mb have oc settings in the mb bios?

    I've been out of ocing and forgotten all the damn setting you need to make just started messing with mine but out of state so I can't help much right now...
     
    trt740 says thanks.
    10 Year Member at TPU
  5. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    Yes I used the settings in the guide. I posted them
     
    10 Year Member at TPU
  6. DOM

    DOM

    Joined:
    May 30, 2006
    Messages:
    7,626 (1.83/day)
    Thanks Received:
    863
    Location:
    TX, USA
    What are the temps like?

    I know when I had the 3770k some didn't oc that good they liked the cold but the ones that did bad in regular cooling didn't do to good on cold

    But 1.35c for just 4.5 seems lil high from 4.4Ghz

    Have your tried turning off HT?
     
    trt740 says thanks.
    10 Year Member at TPU
  7. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    Temps are fine up tell 1.36 then hot hot. Tried turning off the HT and it does not seem to matter there is just a very big voltage jump needed from 4.4ghz and 4.5ghz. However after that it will boot at 4.7ghz . I think this chip is just very average.
     
    Last edited: Jan 16, 2017
    10 Year Member at TPU
  8. exodusprime1337

    exodusprime1337

    Joined:
    Nov 27, 2007
    Messages:
    2,223 (0.61/day)
    Thanks Received:
    375
    honestly had to delid my 3770k to get anywhere.. I've been on a 4790k now for a little over a year holding steady at 4.8 on water temps under 65 full load. My 3770k would hit 80's at 1.25 4.5ghz and after the delid hit 4.8@1.31 hitting 65-70 full stress for heat.

    It's my opinion the 3770k had the worst TIM material ever, or wasn't seated right.
     
    trt740 says thanks.
  9. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    Well I bet on water this would be a good chip. However, It could be ram or something else. I have found certain chips like certain ram better and sometime it is the motherboard. However, it could just be this is a average joe chip. It is just one of the 4 cores failing. Also, it appears that I am going to get 4.4 at 1.25v so thats better not much but still better. Also, I'm not sure I want to delid it because I am not sure how I am going to keep it. I do not want to kill its resale.
     
    Last edited: Jan 16, 2017
    10 Year Member at TPU
  10. Kursah

    Kursah Moderator Staff Member

    Joined:
    Oct 15, 2006
    Messages:
    10,847 (2.69/day)
    Thanks Received:
    5,024
    Location:
    Missoula, MT, USA
    I had good temps for air on my 4790K @ 4.8 on air. But I just went through the delid and reglue procedure.

    I used CLU for the die to IHS. And a thin layer of permatex RTV to glue the IHS back on. Used the socket retention to lock the IHS in-place while everything setup.

    Permatex is good to 170C or something crazy so I'm not concerned. My chip had good contact patterns before the delid, IMHO, and good temps already. But now I'm seeing 10 to 15C cooler and that makes me very happy. I was able to get my system stable at that OC with DDR3 2400. I'm prepping the screenshots for my project thread.

    Frankly if you reglue the IHS back down you'll probably have no problem selling it...hell even if not. Some folks would rather buy a delidded chip than to do it themselves. Maybe of you don't wanna do it but are interested in trying your luck with a delidded chip you could trade someone?

    I purchased the CLU from a seller on eBay for like $10. I barely used any...could probably do another 7-10 more applications. The stuff spreads forever. I warmed mine up first which helped me spread it out. I also used clear nail polish for the strip of resistors next to the die. But I didn't use enough CLU to run down...better safe than sorry.

    I'm debating trying CLU on my 980Ti as the cooler is nickel coated. Gallium (CLU and CLP) vs copper and aluminum = no bueno fwir. I did just reseat with MX4 which brought temps down by 5c so I'm pretty happy anyways.

    Dropping your temps might get you a little further. Might not. Its not bad...I did the razor blade method...will likely do the vice method on my server chip next. Take your time and you'll be fine.

    Otherwise sounds like you're hitting some pretty hard walls for an OC.
     
    trt740 and rtwjunkie say thanks.
    10 Year Member at TPU
  11. trt740

    trt740

    Joined:
    May 12, 2006
    Messages:
    11,069 (2.65/day)
    Thanks Received:
    1,184
    Well to tell the truth 4.4 to 4.6 performance has very little impact but its just the thought.
     
    10 Year Member at TPU
  12. exodusprime1337

    exodusprime1337

    Joined:
    Nov 27, 2007
    Messages:
    2,223 (0.61/day)
    Thanks Received:
    375
    Perhaps, but if he delids and shaves off 10-15c that might be 4.4-4.8 or higher!!

    I say if you aren't terribly worried about the risks, do a delid, the 3770k TIM sucks.
     
  13. Knoxx29

    Knoxx29

    Joined:
    Feb 19, 2014
    Messages:
    4,722 (3.52/day)
    Thanks Received:
    3,341
    Location:
    Behind a VPN
    This was mine.

    1.39V just when tested with LinX, OCCT or Prime95, 1.36V when playing or doing normal tasks.

    downloadfile.jpg
     
  14. eidairaman1

    eidairaman1 The Exiled Airman

    Joined:
    Jul 2, 2007
    Messages:
    18,572 (4.93/day)
    Thanks Received:
    4,543
    I'd say 4.3-4.5
     
    10 Year Member at TPU
  15. Knoxx29

    Knoxx29

    Joined:
    Feb 19, 2014
    Messages:
    4,722 (3.52/day)
    Thanks Received:
    3,341
    Location:
    Behind a VPN
    Did you read my post in the other Thread?
     
  16. Jetster

    Jetster

    Joined:
    Jan 17, 2010
    Messages:
    9,734 (3.43/day)
    Thanks Received:
    6,339
    Location:
    Oregon
    Mine was 4.5. At 4.6 it was just a little hot
     
  17. Knoxx29

    Knoxx29

    Joined:
    Feb 19, 2014
    Messages:
    4,722 (3.52/day)
    Thanks Received:
    3,341
    Location:
    Behind a VPN
    That's why i delided mine;)
     
  18. Jetster

    Jetster

    Joined:
    Jan 17, 2010
    Messages:
    9,734 (3.43/day)
    Thanks Received:
    6,339
    Location:
    Oregon
    I don't do that. It just feels wrong
     
  19. Knoxx29

    Knoxx29

    Joined:
    Feb 19, 2014
    Messages:
    4,722 (3.52/day)
    Thanks Received:
    3,341
    Location:
    Behind a VPN
    Why?

    Note: we do everything to drop our CPU's temps:rolleyes:
     
  20. NdMk2o1o

    NdMk2o1o

    Joined:
    Apr 21, 2010
    Messages:
    4,618 (1.68/day)
    Thanks Received:
    2,034
    Location:
    Redditch, Worcestershire, England
    I think my board really restricts my 3570k as its a low end z77 board and I can only get 4.3 at stock vcore with an + offset of 0.050 which gives me about 1.3 vcore regardless of how much more voltage I put through the chip, frustrating as hell but there's no point in me trying to upgrade the motherboard as I'm hoping to be able to upgrade to Zen this year..
     
  21. NinkobEi

    NinkobEi

    Joined:
    Nov 27, 2006
    Messages:
    2,106 (0.53/day)
    Thanks Received:
    375
    3570k here seems pretty stable at 4.3 w/ 1.25v, so we're in similar ballparks. Oddly enough I find that if I push more voltage through this chip it actually becomes less stable (1.3v @ 4.4 is a no-go). I wonder if that's just a bad analysis on my part
     
    10 Year Member at TPU
  22. basco

    Joined:
    Jun 29, 2009
    Messages:
    774 (0.25/day)
    Thanks Received:
    404
    Location:
    austria
    some chips need the internal pll overvoltage to be enabled at some ghz.
    and vcore overcurrent protection to 120% or more. or called cpu current capabillity.
    northbridge frenzy =every chip is different and some need to be 400 to 500 lower.
    cpu input voltage- i have mine on 1,640-but more can help too depends on the chip
     
    Last edited: Jan 16, 2017
  23. Sasqui

    Sasqui

    Joined:
    Dec 6, 2005
    Messages:
    9,564 (2.20/day)
    Thanks Received:
    3,593
    Location:
    Manchester, NH
    I'm at 4.7 on my delidded 3570k and its been flawless for a good 4 yrs. Gets up to 90c under stress and I don't even know what vCore is, I'd have to check. The IMC can affect stability too, that may be what's going unstable with more voltage in your case, not sure.
     
    10 Year Member at TPU
  24. Recon-UK

    Recon-UK

    Joined:
    Jun 21, 2016
    Messages:
    1,894 (3.87/day)
    Thanks Received:
    1,710
    I remember back on the 2500k that hitting 5ghz and beyond called for very strict PLL tweaks, and other unusual tweaks like I/O voltage, and HDD's would have issues loading up and things not keeping up with the CPU.

    It get's very hard to push that kind of speed on Sandy at least, and Ivy being hotter...
     
  25. NinkobEi

    NinkobEi

    Joined:
    Nov 27, 2006
    Messages:
    2,106 (0.53/day)
    Thanks Received:
    375
    Right now my ram is 1600 @ 1.65v, so you are saying I might have better luck with lower speed/voltage memory?
     
    10 Year Member at TPU

Currently Active Users Viewing This Thread: 2 (0 members and 2 guests)