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

e8500 e0: getting over 4ghz without killing it

Discussion in 'Overclocking & Cooling' started by Mr.Amateur, Mar 1, 2009.

  1. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    As the title says, I have issues with my particular system...

    e8500 e0 @ 4.00Ghz (500x8) 1.34V
    Asus P5q Pro p45 chipset 1613m bios
    Patriot Extreme Performance 2x2 GB DDR2-6400 (overclocked with cpu) 5-5-5-15 1:1
    *see system specs for power supply and other information)

    dRAM voltages are set at 2.1V
    NB at 1.26V
    SB at 1.1V
    FSB termination voltage at 1.26V as well
    333 FSB strap

    I am at my wit's end trying to get past 4.0xx, but no matter what I try further overclocking fails. I almost passed an occt 2 hr stress test running at 4250mhz (1.4 core voltage), but 1hr 30mins in I am greeted with a black screen and restart. No matter how much I increased vcore, up to 1.45V, my overclocking failed. Though 4Ghz is not a miserly speed, I have seen other e8500 e0 overclocks on the internet faster with less voltage! I am about to replace my memory with a 2x2GB OCZ reaper set (1066), so I wonder...

    1. Is RAM holding me back?
    2. Since the memory has not arrived yet, is it still possible to obtain a stable (2hr occt) 4.2 ghz overclock below 1.4V? If so, any advice on how to do it?

    Sorry for being wordy, it's the passion for overclocking :banghead:
     
  2. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    thanks for replying :)
     
  3. thebeephaha

    thebeephaha

    Joined:
    Apr 28, 2007
    Messages:
    1,182 (0.41/day)
    Thanks Received:
    104
    Location:
    Bellevue, WA
    I think your NB and SB need more.
     
  4. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    SB I can understand, but I don't have active cooling for my NB, so that's capped at 1.3V
     
  5. DonInKansas

    DonInKansas

    Joined:
    Jun 2, 2007
    Messages:
    5,104 (1.80/day)
    Thanks Received:
    1,275
    Location:
    Kansas
    NB capped at 1.3? I run my p5k at 1.55 with no issues on stock cooling. I don't even think 1.3 is an option.:p

    Asus boards like the voltage.

    Have you tested the RAM? Maybe they are bunk. (Of course the new Reapers will fix that issue.:p )
     
  6. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    Well I tried imputing 1.5V on my NB, but the numbers turned purple:ohwell:
    So instead I set NB and vcore to 1.4V (450x9.5). Got to desktop alright, but restarted after 2 seconds of OCCT testing:( No idea how you're using 1.55 on the NB Don :confused:
     
  7. Kursah

    Kursah

    Joined:
    Oct 15, 2006
    Messages:
    8,122 (2.66/day)
    Thanks Received:
    1,884
    Location:
    Missoula, MT, USA
    The number going purple in bios just is a means for warning, just keep an eye on your NB temps using the provided MB monitorring software or a program like Everest, I've ran my P5Q Deluxe NB up to 1.55v before, and temps didn't go over 41C.

    Another thing I run my FSB and NB at similar values, for me some reason this seems to be some kind of voodoo OC stability helper in my mind, but when OC-ing my e8600 to 4.5Ghz and e8500 to 3.8-4.0Ghz it seemed to help.

    Though, even for 500FSB I didn't need more than 1.26v on either the NB or FSB voltages. As far as getting further, it could be a limitation of the chip. What kind of temps are you getting at these attempted overclock runs in the goal of greater than 4GHz? I don't blame you for trying, the e8500 I have is a C0, it needs 1.23v for 3.8Ghz and 1.29v for 4.0Ghz, but from there the voltage requirements shoot up quite a bit, I needed around 1.45v for 4.4ghz iirc. Not worth it, the chip's sensors read cool.

    What bios version are you running? Is it a modded bios? There might be something there, doubt it, but doesn't hurt to ask!

    I'm going to have to say there's some sort of limitation, whether it's board or chip, it's there, could you list all of your OC settings? Also try an FSB strap set to AUTO, I've had better luck leaving it at Auto than manually setting it.

    :toast:
     
    Mr.Amateur says thanks.
  8. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    It's possible to monitor NB temps with everest?!:confused: If that's true, where to find such utility?

    My bios is a modded one: 1613m (made by Ket)

    For temps, 36/39 idle cores, and 60/62 on 100% load (Freezer 7 pro is awesome:))

    I'll try changing the FSB strap to Auto later, but by oc settings do you mean everything that CAN be manipulated, or what I changed from auto? Anyways, with everything else on auto, my most stable OC (4.00 ghz):

    CPU Multi:8
    FSB: 500
    FSB Strap: 333
    PCIE Freq: 100 (no point in changing...)
    DRAM Timings: 5-5-5-15
    DRAM Frequency: DDR2-1002 (likes going a little over)
    CPU Voltage: 1.34V
    FSB Termination: 1.26V
    NB Voltage: 1.26V
    SB Voltage:1.10V
    DRAM Voltage: 2.1 V (playing it safe)
    CPU PLL Voltage (still clueless on what it does): 1.50V

    ...that's all I can remember
     
  9. JrRacinFan

    JrRacinFan Served 5k and counting ...

    Joined:
    Mar 17, 2007
    Messages:
    19,421 (6.68/day)
    Thanks Received:
    4,496
    Location:
    Youngstown, OH
    Last edited: Mar 1, 2009
    Mr.Amateur says thanks.
  10. DonInKansas

    DonInKansas

    Joined:
    Jun 2, 2007
    Messages:
    5,104 (1.80/day)
    Thanks Received:
    1,275
    Location:
    Kansas
    Change your FSB strap to Auto; see if that helps.
     
    Mr.Amateur says thanks.
  11. Kursah

    Kursah

    Joined:
    Oct 15, 2006
    Messages:
    8,122 (2.66/day)
    Thanks Received:
    1,884
    Location:
    Missoula, MT, USA
    Google Everest, its' gotta be a newer version, the beta subscribtion isn't exactly cheap, but Everest is a very kickass utility! Though you could just use the provided Asus PC Probe utility too.

    I no longer use Ket's modded bioses as I found absolutely no difference in overclocking stability or higher overclocks, I'm on version 1702 directly from Asus...I dunno what version number that would be for the Pro's though, and really stability and OC-ability didn't seem to change from version 1306 from my perspective.

    I still say try FSB Strab at AUTO, though it might not change anything.

    So you have the CPU Load-Line Calibration enabled? If not, you shold, that will help provide more solid voltage values to the CPU. I would say try your SB voltage at 1.20v. Change CPU PLL voltage to 1.54-1.56v...seemed to help my e8600 with 4.5Ghz stability.

    You would only need 444FSB to attain around 4.2GHz at the full 9.5X multi, that might help with the memory OC a little, and would allow you to more focus on your CPU OC at this time. Set the memory divider to a 1:1 and get DDR888 speeds for now, definately some more breathing room for now.

    Also make sure that the Spread Spectrum options stay disabled, on my board they are default disabled, so that shouldn't be an issue with yours either iirc. Keep us posted! It looks like you're setting things right, but question is 1.34v for 4Ghz? What does it read in windows idle/load? I use Everest, PC Probe or CPU-z for that value while running orthos in small FFT's.

    :toast:
     
    Mr.Amateur says thanks.
  12. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    Wow...1.5V for 4.5Ghz:twitch:
    8.5x470 is stable, then again most ocs about 4ghz on this are stable...
    Changed NB to 1.56V and fsb strap to auto, gonna try that with a higher oc

    Ah, Load Line Calibration is disabled; had some stability issues when it was on. I'll try that again..
    As for bios, there aren't many changes from the ones on ASUS's website.
    1.56v on CPU PLL...interesting
    As for memory dividers, P5Q Pro has nothing that allows me to change that, so as far as I know it's ALWAYS 1:! :p I'll try 444x9.5 without changing vcore...
    Spread spectrum is disabled, 1.34 in the BIOS, which translates to max. of 1.33V on cpu-z (idle 1.32V)

    Thanks everyone :) (doesn't mean the end yet...)
     
  13. Kursah

    Kursah

    Joined:
    Oct 15, 2006
    Messages:
    8,122 (2.66/day)
    Thanks Received:
    1,884
    Location:
    Missoula, MT, USA
    The memory dividers aren't shown in the x:x format, but rather the speed of your memory at a certain ratio. I.e. at 400FSB you would be 1:1 at 800 speed, 4:5 at DDR1000 and so on.

    LLC should help with more stable voltages, I've not heard of instability from it, but it should help you lower your CPU voltage values closer to what the chip actually needs with lower vdroop.

    Keep us posted on your results! :toast:
     
    Mr.Amateur says thanks.
  14. r9

    r9

    Joined:
    Jul 28, 2008
    Messages:
    2,144 (0.89/day)
    Thanks Received:
    284
    Increase NB SB voltage. Try other CPU multi. Lower memory speed just to rule it out.
    Don`t compare your cpu with other not all are the same. And 1.5 hours is stable enough for me.
     
    Mr.Amateur says thanks.
  15. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    Well, as I tried to go 9.5x444, my computer went haywire :(; I'm running it at stock speeds...for now... I assume my NB had something to do with my malfunction

    Must stay vigilant!
     
  16. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    So far so good, 444x9.5 at 1.36V in the BIOS (1.344V on cpu-z)

    Unfortunately, I fear this stability will not last after 1 second on OCCT...well, time to find out.

    EDIT: Comp froze on me while trying to reply; also 1.5 on NB makes restarting impossible (on...off) Error was detected on core 0 after 2 seconds of OCCT...overclocking is a serious business:banghead:
     
    Last edited: Mar 1, 2009
  17. JrRacinFan

    JrRacinFan Served 5k and counting ...

    Joined:
    Mar 17, 2007
    Messages:
    19,421 (6.68/day)
    Thanks Received:
    4,496
    Location:
    Youngstown, OH
    Remember though that was a C0 and only ran it for a few moments. I wouldn't dare try to stable it out for 24/7.
     
  18. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    I wonder...is it possible to avoid increasing NB voltages by replacing RAM? If not, then I'll invest in a 40mm fan.
     
    Last edited: Mar 1, 2009
  19. pabloc74

    pabloc74 New Member

    Joined:
    Feb 11, 2008
    Messages:
    184 (0.07/day)
    Thanks Received:
    11
    Location:
    Argentina
    trying touch only cpu volt and mem volt, try to set the mem the loweest as you can, and then try max oc.
    i don't know much about oc, but with only touch vcore i make this..

    [​IMG]
     
    Mr.Amateur says thanks.
  20. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    Woah, is that fully stable pabloc? Also, you do mean touch cpu voltages and memory voltages right, and set the lowest volts for memory? Or am I misinterpreting something..
     
  21. pabloc74

    pabloc74 New Member

    Joined:
    Feb 11, 2008
    Messages:
    184 (0.07/day)
    Thanks Received:
    11
    Location:
    Argentina
    in that case the vcore was 1.35 and memory was reaper 1150 2.1v (stock) (1100mhz on screen with 2.1v) stable 24/7:), and see taht was an C0 step, your cpu could be better
     
    Mr.Amateur says thanks.
  22. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    Amazing! :rockout: Looks like upgrading to those reapers will be worth it.

    For the time being (before my memory arrives from newegg), I'll be tinkering with the bios to get some results.

    Note: I switched the position on my Freezer 7 Pro to point towards the NB. It looks strange, but it should draw heat away from the NB heatsink.
     
  23. Scrizz

    Scrizz

    Joined:
    Aug 22, 2007
    Messages:
    2,989 (1.09/day)
    Thanks Received:
    416
    Location:
    Florida, US
    definately try setting you mem lower
     
  24. Mr.Amateur New Member

    Joined:
    Jan 19, 2009
    Messages:
    188 (0.08/day)
    Thanks Received:
    10
    By mem, do you mean timings?

    Edit: NB HAET 1.5V! hopefully new memory will create a work-around.
     
    Last edited: Mar 2, 2009
  25. pabloc74

    pabloc74 New Member

    Joined:
    Feb 11, 2008
    Messages:
    184 (0.07/day)
    Thanks Received:
    11
    Location:
    Argentina
    no, frequency:)
     
    Last edited: Mar 3, 2009
    Scrizz and Mr.Amateur say thanks.

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

Share This Page