Question on my Opteron 165

edited May 2006 in Hardware
Hi all
Have been doing some overclocking to my opt165
Things seem really stable & pretty uneventfull up to around 279x9 @ 1.375V
After that It seems that I have to raise the volts at a pretty steep curve to get to to 300+ Prime stable. Have spent the last couple of weeks trying to cut back on the volts. Not much success.I am on water and heat is not a issue (104F tops)


I guess I have 2 questions

Is that the nature of some (not so good) Opte165s ?

Could it be something in the motherboard causing this. The board always seems to want to overvolt above whatever settings. In the dashboard and Bios at what ever speed, always to the plus side +.2 to .4 volts more up and down up and down.

Or is al this normal normal
Just wondering?

Comments

  • lemonlimelemonlime Canada Member
    edited May 2006
    Welcome to Short-Media, Nobody! :cheers:

    Thats pretty much the nature of just about all K8 processors I'm afraid. Most of the chips I've seen behave in that manner. You'll usually go from close to default vcore to 1.6V just to obtain an exta 100-150MHz.

    There are a few tips I can offer:
    -Be sure the HTT multiplier set to 3X at that reference clock.
    -Ensure your memory is on a divider (at least to rule it out).. Keep it as close to 200MHz as possible for initial testing.
    -Keep the power regulation circutry on the board cool (mount an 80mm fan if possible)
    -Increase chipset voltage by 0.1V if possible.

    Could you provide some more info on your setup?
  • edited May 2006
    Thanks for the welcome and info
    Was Just wondering about the voltage thing.
    I have another question - you see around opte165s that do 3000Mhz @1.5V
    Was kind of wondering - given the same CPU and If you had some sort of extream cooling on that CPU if it would run with less volts than on the same chip say 30 degrees warmer?


    Opte165
    Abit AT832
    OCZ rev 2 2225 ram
    A Mad Dog 650 power supply
    X1900 xt
  • edited May 2006
    Yes, that is quite possible, using a phase change setup Like a Vapo LS.

    And FYI, both AMD's and Intel's processors or late act this very same way. Once you have to start jacking up the vcore significantly from stock vcore, you won't see much more speed for large increases in vcore. I guess this is just a byproduct of the 90 and 65 nm manufacturing processes both are using now.
  • lemonlimelemonlime Canada Member
    edited May 2006
    Nobody wrote:
    Thanks for the welcome and info
    Was Just wondering about the voltage thing.
    I have another question - you see around opte165s that do 3000Mhz @1.5V
    Was kind of wondering - given the same CPU and If you had some sort of extream cooling on that CPU if it would run with less volts than on the same chip say 30 degrees warmer?

    Yep, absolutely. There is a generally accepted rule that for every 10'C decrease in temperature, an integrated circuit (like a CPU) should be able to increase it's operating frequency by 3%. Silicon simply works more effectively at lower temperatures.

    Most dual core processors in the 3GHz range are phase-change cooled, although there are a few rare exceptions.

    Here is a graph I made in my Vapochill LS review:

    VAPO_PERFORMANCE1.JPG

    As you can see, just by lowering the temperatures, I was able to increase my overclock by about 400MHz.

    EDIT: Replied at about the same time as Mudd :)
  • edited May 2006
    Yeah, but I didn't post a purty graph like you, lemonlime. ;)
  • edited May 2006
    Thanks for the info
Sign In or Register to comment.