Sign in with
Sign up | Sign in
Your question
Closed

Optimize i5-3570K overclock

Last response: in Overclocking
Share
a b à CPUs
January 17, 2013 1:43:44 AM

Looking to optimize my overlock to hit 4.8 GHz. If that's simply not do-able on my particular processor, I'd like some tips on how to optimize (lower Vcore) my current 4.7 GHz overclock.

This particular OC won't be 24/7 - even if I get voltages down to a reasonable level, I don't have a practical use for it. This is mostly just to see how far I can reasonably push my particular 3570K and gain experience overclocking in the meantime.

I figured I'd try to give you guys as much info as possible, so here goes:

Relevant components:
  • CPU: i5-3570K, currently stable at 4.7 GHz (more details below)
  • Motherboard: ASUS P8Z77-V LK
  • Cooler: Cooler Master Hyper 212+

    Settings I've tweaked to get this OC so far:
  • Multiplier set to x 47
  • Vcore set manually to 1.385V
  • Load-line Calibration set to Ultra High (ASUS setting corresponds to 4 on other boards, I think)
  • CPU Current Capability set to 120% (not sure how effective this is, if at all. Still testing.)
  • CPU Voltage Frequency set to 350KHz (again, still testing for effectiveness)
  • CPU Power Phase Conrol set to Extreme

    Tests
  • Several 1 hour runs of 10 passes of IntelBurnTest on maximum (I usually manage to run it using ~6GB of RAM)
  • Haven't done an extended run of Prime95 on this particular configuration yet.
  • Downloaded OCCT, but haven't read up or used it yet either.

    Stats while testing:
  • Max CPU temp never touches or goes past 50C
  • Vcore at load averages 1.39V, touched 1.41V briefly at some point

    Finally, my question:
    Is there anything else I can optimize to potentially hit 4.8 GHz? If not, is there anything I could do to lower Vcore at 4.7 GHz + power consumption reductions? Is there anything wrong or potentially risky about my current settings?
    a b à CPUs
    a b K Overclocking
    January 17, 2013 2:15:46 AM

    What are you using to test your temperatures? 4.6GHZ @ 1.40V does not make sense that you are only hitting 50C, unless your processor is throttling massively
    Score
    0
    a b à CPUs
    January 17, 2013 2:29:47 AM

    Adroid said:
    What are you using to test your temperatures? 4.6GHZ @ 1.40V does not make sense that you are only hitting 50C, unless your processor is throttling massively


    I am facepalming myself so hard right now. I've been using HWMonitor. Now that I think about it, I had the exact same issue of false temps on my last build and got Core Temp instead, yet somehow completely forgot about it this time around.

    Just downloaded Core Temp again, and realized I was dangerously close to Tj. Max for the hours I've been running IBT... Temps broke 90 within seconds of running IBT. :sweat: 

    Hopefully I haven't done too much lasting damage...
    Score
    0
    Related resources
    a b à CPUs
    a b K Overclocking
    January 18, 2013 12:39:15 AM

    I was wondering the same thing too when I saw the "max temp never touches 50c" at that clock speed and voltage makes sense now though.

    For the 4.8 ghz overclock you are looking for it's not a question if it's doable with the particuar procssor you have, the obstale is the cooling you are using - that Hyper 212 is being pushed above it's limits in my opinion, the max for a budget cooler like that is 4.2-4.4ghz, you are going to have to looking into water cooling to keep a leash on temperatures for the goal you are seeking to accomplish.

    Score
    0

    Best solution

    a b à CPUs
    a b K Overclocking
    January 18, 2013 12:57:50 AM

    Yikes! Well the good news is that you somehow got through 10 runs of IBT with your Ivy Bridge CPU at 4.7ghz with a Hyper 212+. Usually Ivy gets unstable and crashes at the temps you were getting. That's somewhat remarkable, but very bad for the longevity of your chip. Since it survived, it'll probably be fine but make sure to read at least two Ivy/Sandy bridge guides for overclocking beforehand. You'll learn that 50c @ 1.4v under load isn't possible on air for an Ivy.

    Also - on a new CPU I always try multiple CPU temp monitoring programs until I'm sure that a particular program is reading accurately. Anyway, here's an article I read when messing around with Ivy when it launched: http://www.overclock.net/t/1247413/ivy-bridge-overclock...

    And here's another one: http://www.anandtech.com/show/5763/undervolting-and-ove...
    Share
    a b à CPUs
    a b K Overclocking
    January 18, 2013 10:08:48 AM

    As others have posted above, I wouldn't expect more than 4.2-4.4 ghz from a 212.

    When using an air cooler for ivy bridge, 99% of the time temperatures will be your limiting factor on the OC, not voltages.

    For a point of reference, my 3570K with CM 212 Evo is completely stable @4.2ghz with 1.235 V, completely stable @ 4.3GHZ with 1.240 V.

    But, at 4.4GHZ, 1.265 V was still giving annoying WHEA errors in the event logger, not to mention hitting mid 80C during prime95 large FFTs

    Now I am sitting happily at 4.3GHZ @ 1.240 V. I didn't win the "silicone lottery" with my chip, but I am fast and stable at temperatures I am comfortable with, and hence the stopping point of my overclock.
    Score
    0
    a b à CPUs
    January 19, 2013 3:20:27 PM

    Thanks for the replies, I knew something was too good to be true. I've since dropped my 24/7 OC to a more modest 4.1 GHz at stock Vcore. I'll do some more research and see if I can't tweak some more while keeping things cool.

    Heading off to college in a few months, so I'm probably going to hold off on investigating water cooling until I get settled down there.
    Score
    0
    a b à CPUs
    a b K Overclocking
    January 20, 2013 1:06:32 PM

    Adroid said:
    As others have posted above, I wouldn't expect more than 4.2-4.4 ghz from a 212.

    When using an air cooler for ivy bridge, 99% of the time temperatures will be your limiting factor on the OC, not voltages.

    For a point of reference, my 3570K with CM 212 Evo is completely stable @4.2ghz with 1.235 V, completely stable @ 4.3GHZ with 1.240 V.

    But, at 4.4GHZ, 1.265 V was still giving annoying WHEA errors in the event logger, not to mention hitting mid 80C during prime95 large FFTs

    Now I am sitting happily at 4.3GHZ @ 1.240 V. I didn't win the "silicone lottery" with my chip, but I am fast and stable at temperatures I am comfortable with, and hence the stopping point of my overclock.


    You need 1.235v for a 4.2ghz overclock to run stable ? That just seems very high to me even if you didn't win the silicone lottery.
    Score
    0
    a b à CPUs
    a b K Overclocking
    January 20, 2013 10:38:07 PM

    Orlean said:
    You need 1.235v for a 4.2ghz overclock to run stable ? That just seems very high to me even if you didn't win the silicone lottery.


    I may have been able to do slightly less for 4.2GHZ, but I wanted to get faster speeds so I didn't stay there long.

    Are you familiar with WHEA errors?
    Score
    0
    a b à CPUs
    January 31, 2013 12:36:52 AM

    Best answer selected by xomm.
    Score
    0
    a b à CPUs
    a b K Overclocking
    January 31, 2013 1:55:02 AM

    ok thought i may interject...

    I just got my 3570k, to oc at 4.7ghz, with a 212 evo, on 1.296v.

    am i missing something? as pretty much every guide i read says, 85c is a good max temp. 95c for stress testing. my highest temp thus far has been 88c. 90c at one point but i have made some adjustments.
    Score
    0
    !