The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.

    G51VX undervolting inconsistencies

    Discussion in 'Asus' started by jenesuispasbavard, Oct 19, 2009.

  1. jenesuispasbavard

    jenesuispasbavard Notebook Evangelist

    Reputations:
    51
    Messages:
    412
    Likes Received:
    2
    Trophy Points:
    31
    So, I got my Asus G51VX-RX05 from Best Buy this week and immediately overclocked the P7350 to 2.5 GHz from the stock 2.0 GHz (one step at a time of course). The FSB is now rated at a cool 1333 MHz, from 1066 MHz.

    I decided to undervolt after I noticed the temperatures were getting a bit high. The optimization guide tells you to use CrystalCPUID to undervolt and CPU-Z to check the voltage.

    Here's the problem - stock voltage according to CPU-Z is 1.138V (at multiplier 7.5x), but in CrystalCPUID it's 1.25V (also in the optimization guide). When I undervolt to 1.2V in CrystalCPUID, the voltage in CPU-Z goes down to 1.088V, and the guide tells you not to go below 1.113V in CrystalCPUID. Also, Everest tells me the voltage is the same as the one CPU-Z reports.

    According to Intel's P7350 page, the VID voltage range is 1.0-1.25V.

    So, which one should I believe, CPU-Z or CrystalCPUID?
     
  2. Abula

    Abula Puro Chapin

    Reputations:
    1,115
    Messages:
    3,252
    Likes Received:
    13
    Trophy Points:
    106
    Well i havent tested undervolting on cpu yet, but your best bet probably is asking inside Tev's thread as most that have done it will see your post there, or PM Tev, since he made the guide and has the same processor.

    But there are also other guides that might be helpful, like The "Undervolting" Guide

    Good luck,
     
  3. ClearSkies

    ClearSkies Well no, I'm still here..

    Reputations:
    1,059
    Messages:
    2,633
    Likes Received:
    0
    Trophy Points:
    55
    It's generally a poor idea to try and undervolt a cpu after it's been overclocked, as the overclock usually requires the extra voltage buffer that Intel builds into their chips (i.e. stock voltage is always higher than necessary, which is why you can undervolt a cpu at stock clocks). This is a common issue in extreme desktop overclocks, that you have to apply extra voltage in order to help the cpu overclock hold.

    Undervolting an overclocked cpu is a path to blue-screen purgatory :).
     
  4. dancom96

    dancom96 Notebook Consultant

    Reputations:
    11
    Messages:
    211
    Likes Received:
    0
    Trophy Points:
    30
    Believe CrystalCPU and Intel's site. CPU-z always shows a bit higher voltage than it really is.

    I got my P7450 overclocked to 2.55ghz from 2.13ghz, and undervolted to 1.075v from 1.1375v

    *edit*
    By the way how'd you get such a high OC?
     
  5. jenesuispasbavard

    jenesuispasbavard Notebook Evangelist

    Reputations:
    51
    Messages:
    412
    Likes Received:
    2
    Trophy Points:
    31
    Yes, I know it's a poor idea, so I undervolted after downclocking back to 2.0 GHz and went up from there, again to 2.5 GHz. It's stable with Prime95 with a full cycle of 8K FFTs and another full cycle with 10K FFTs (I know, I should test it for longer, but I wanted to make sure the voltage wasn't too low). Also 1024K FFTs work fine for ~30 minutes (I didn't test longer), so the RAM is fine at higher FSB as well.

    And it's not much of an undervolt, if CrystalCPUID is to be believed, only 0.05V.

    I've also been running SETI@Home for five days now and haven't had any computation errors.

    But my CPU-Z shows a bit lower voltage than it actually is... CPU-Z is showing 1.138V when CrystalCPUID says 1.25V (at stock voltage, multiplier 7.5x), and 1.088V when CrystalCPUID shows 1.20V (undervolted).

    Intel's site shows that the max voltage (I assume it is stock, and at multiplier 7.5x) is 1.25V, which is what CrystalCPUID tells me at stock voltage.

    The OC is not too high, only 80MHz more than yours. And you've underclocked 0.0125V more than me.

    BTW, what's your voltage at zero CPU usage (when it clocks down to a 6.0x multiplier)? They should be the same in both our cases since I haven't changed that voltage yet. For me, it's 0.925 according to CPU-z, 0.93V according to Everest, and 1.113V according to CrystalCPUID.

    Bah.. too many numbers...
     
  6. dancom96

    dancom96 Notebook Consultant

    Reputations:
    11
    Messages:
    211
    Likes Received:
    0
    Trophy Points:
    30
    Ah didn't notice that part. Just ignore what CPUz said and go with CrystalCPUID. Don't know about Crystal, (I use RMClock), but you should be able to go under 1.113 if it's stable.

    Yours is high compared to mine since you start off 130mhz less than me, and end up 80mhz more. I can't overclock any more even after I put it back to stock volts.
     
  7. jenesuispasbavard

    jenesuispasbavard Notebook Evangelist

    Reputations:
    51
    Messages:
    412
    Likes Received:
    2
    Trophy Points:
    31
    Thanks for the info, I'll take CrystalCPUID's readings as accurate from now on.

    Back to you vs me, I said my total overclock (2.5-2.0GHz = 500 MHz OC) was only 80MHz more than your total (2.55-2.13GHz = 420 MHz OC). Not that big a deal, considering I've only dropped my voltage by 0.05V (you've dropped yours by 0.0625V). And the P7350 is probably just a factory-underclocked (or under-multipliered) P7450, or vice versa.