Ivy Bridge Memory Voltages

Discussion in 'CPUs and Overclocking' started by kithylin, Jul 12, 2012.

  1. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    This is somewhat overclocking related and more of general system related.

    I'm upgrading from an I7-920 LGA-1366 system to an Ivy Bridge core i5 system and I'm wondering about memory voltages.

    I know with the i7-920 we could set the memory to a maximum of +0.50v from the cpu's core voltage.

    Is this still true with ivy bridge, or does the cpu and memory have to be of the same voltage now?
     
  2. ShintaiDK

    ShintaiDK Lifer

    Joined:
    Apr 22, 2012
    Messages:
    20,091
    Likes Received:
    15
    Max is 1.575V on Ivy before warranty void.

    On your i7 920 it was 1.65V.
     
  3. IntelEnthusiast

    IntelEnthusiast Intel Representative

    Joined:
    Feb 10, 2011
    Messages:
    582
    Likes Received:
    0
    +1 Thanks ShinaiDK

    The voltage on the memory for the 2nd and 3rd generation Intel® Core™ processors is 1.5v (1.425v to 1.575v). Of course if you buy the Intel® Performance Tuning Plan you can run whatever you want without worrying about voiding your warranty.
     
  4. pantsaregood

    pantsaregood Senior member

    Joined:
    Feb 13, 2011
    Messages:
    937
    Likes Received:
    0
    Running lower voltages don't void your warranty, do they? I think SB/IB support DDR3L.
     
  5. Ferzerp

    Ferzerp Diamond Member

    Joined:
    Oct 12, 1999
    Messages:
    6,107
    Likes Received:
    1

    Outside of spec is outside of spec. There may be areas where relative voltages matter, and lowering one voltage creates a larger differential than otherwise would exist, so yes, it is potentially the same problem.

    If they suppory low voltage explicity, than that is in the ok range, so don't worry about it. If they don't, don't assume it is ok to do it though. The risk is low, but technically, you're still ruining your warranty.
     
  6. AsusGuy

    AsusGuy Senior member

    Joined:
    Dec 9, 2004
    Messages:
    228
    Likes Received:
    0
    I didn't know about the Performance Tuning Plan. Seems like an interesting idea.
     
  7. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    Well the thing is the voltage differential on the i7 is 0.50v, which means if the cpu was set to 1.35v or 1.45v, then the ram could be 1.85v, or 1.95v without damaging the CPU. This isn't well known or well documented but it is true and does work.

    I'm just wondering if the same is true with ivy bridge, or is it a hard 1.575v max for ivy bridge flat out?
     
    #7 kithylin, Jul 12, 2012
    Last edited: Jul 12, 2012
  8. Ferzerp

    Ferzerp Diamond Member

    Joined:
    Oct 12, 1999
    Messages:
    6,107
    Likes Received:
    1

    That was myth by the way, IIRC.
     
  9. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    No.. It's actually true, it's reported by many overclockers on many differnet forums, and I've used it myself on my i7 for 2.75 years and it worked perfectly fine with no issues.

    This thread is not about speculating about the i7's voltage systems, it's to confirm or deny if ivy bridge works the same way.
     
  10. Ferzerp

    Ferzerp Diamond Member

    Joined:
    Oct 12, 1999
    Messages:
    6,107
    Likes Received:
    1

    I don't think you understand how one proves a concept.

    Your assertion: Memory voltage must remain within 0.5V for safe operation.

    Your "proof": I've run memory within 0.5V and nothing has blown up.

    The problem: If memory doesn't need to be within 0.5V to remain safe, it still wouldn't blow up if within 0.5V.

    You can only prove it by consistently breaking processors with values not within 0.5V. Not by not breaking processors within 0.5V.

    The statement you must prove is the following: Running these two voltages greater than 0.5V apart will damage processors.
     
  11. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    Thank you for de-railing my thread and running it off topic, again, I'm not trying to discuss I7 here, this thread is for IVY BRIDGE If you don't have anything useful to say about Ivy Bridge voltage systems, please keep your comments to yourself.
     
  12. Kenmitch

    Kenmitch Diamond Member

    Joined:
    Oct 10, 1999
    Messages:
    6,573
    Likes Received:
    9
    Gave me the courage to push my 2550k to 5.6ghz(56x) just to see....Only did 2 cores just in case :)

    Seems like a decent deal for some just in case protection to me. Some people look at is as keeping a person honest. The not so honest ones look at it as a waste of money. I got my 2550k for cheap with the microcenter facebook coupon so I figured what the hell....Takes a lot to kill a Sandy Bridge it looks like so far. I'm thinking mine would kill my motherboard before it dies an untimely death!
     
  13. Ferzerp

    Ferzerp Diamond Member

    Joined:
    Oct 12, 1999
    Messages:
    6,107
    Likes Received:
    1

    Multiple anectdotes does not become useful statistical data. One cannot "confirm or deny if ivy bridge works the same way" when "the same way" is a specious understanding of the prior generation. Provide some sort of proof for your conjecture, then we can discuss if the subsequent generations behave similarly. Otherwise, it is about as useful as discussing, since the moon is made of cheese, if other celestial bodies are as well.
     
  14. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    I was just hoping originally that someone here that has experience overclocking ivy bridge could come in and confirm or deny having their ram +0.5v or so from the processor works and doesn't damage anything. A simple yes, or no, it's not a long-winded discussion about the topic.
     
  15. utahraptor

    utahraptor Senior member

    Joined:
    Apr 26, 2004
    Messages:
    901
    Likes Received:
    0
    My friend has been running his 3770k at 1.65 memory volts with the 2400 trident 8 gig modules for about 2 months now. It has not fried yet.
     
  16. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    Aha, Thank you. This is the kind of information I was hoping for ;)
     
  17. ShintaiDK

    ShintaiDK Lifer

    Joined:
    Apr 22, 2012
    Messages:
    20,091
    Likes Received:
    15
    Just because it doesnt fry in 2 months doesnt mean its good. Overvoltage of the memory gives a possible funny sudden death now and then. The memory controller in the CPU simply stops working. Its seen with Nehalem and onwards.
     
  18. cmdrdredd

    cmdrdredd Lifer

    Joined:
    Dec 12, 2001
    Messages:
    24,522
    Likes Received:
    3
    There are many who have been running SB CPUs with 1.65v memory left over from past builds no problem. I don't see why Ivy would be different. I don't like high voltages, I try to stay at or under 1.5v if possible.
     
  19. kithylin

    kithylin Member

    Joined:
    Jan 5, 2010
    Messages:
    130
    Likes Received:
    0
    I'm not trying to run 1.65v memory in my ivy bridge at 1.65v, this kingston HyperX ddr3 I have is rated at 1.65v but seems to be perfectly happy at either 1.5v or 1.65v. I was just wondering about how ivy bridge functions for future knowledge if and when I ever decide to overclock it.
     
  20. utahraptor

    utahraptor Senior member

    Joined:
    Apr 26, 2004
    Messages:
    901
    Likes Received:
    0
    I guess if you want to be 100% confident at higher than 1.5 you should get the optional over locking warranty.