Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Safe vCore Settings

Options
  • 19-05-2004 11:10pm
    #1
    Registered Users Posts: 968 ✭✭✭


    I've got a very stable OC of 3.6GHz on a P43.0C with the vCore set at the minimum for the board, 1.53v (if i recall correctly). Very happy with that.

    But of course it's just never enough! I have tried an FSB of 247, 248 and 250, and the only way i can get it to run for any amount of time at this level is if I up the vCore to 1.7v. ANything lower and the system locks up after running prime for 5-10 mins. I'm currently priming at 247 at 1.7v with the RAM at 5:4, to be sure I am testing the CPU properly, and so far so good.

    I did this a while ago (under the influence :)) and someone commented that I was brave (:p). Is this dodgy? I see in some OC databases that people have all sorts of high voltages, and as the chip says 1.55 max, I'm a little... unsure... :)

    CPU temp is maxing at 56 with this voltage, a good 5 degrees hotter than when the vCore it at its lowest, but I'm still in safe territory.

    Any opinions/advice? Will I end up with a China Syndrome in my case? :)

    "Today, only a handful of people know what it means... Soon you will know."


Comments

  • Registered Users Posts: 2,754 ✭✭✭Col_Loki


    If you had an XP id say up the voltage to 2v and happy hunting!! Its different with the P4's , lower temps & voltage are acceptable. Ive heard of some instant deaths on P4's but this generally with older revisions and 1.85v+ running 24/7 with this setting.

    This question is far better suited to the intel heads that float round these boards so i wont go much further :) . Its nice to be able to have lets say 3.8ghz on high voltage for benchmarks or showing off isint it :) , 3.6v on 1.53v is very nice and not within the slightest bit risky!!

    BTW ive tried overclocking when "under the influence" and some of my best work was done there :) . But i did try my Xp2500+ @ over 3ghz and locked up the whole system (slight misscalculation eh :rolleyes: ).


  • Registered Users Posts: 771 ✭✭✭Sir Random


    I wouldn't run it 24/7 at 1.7v on air, P4s have suffered sudden death from overvolting . Highest I've taken mine was 1.65v (to break 4Ghz).


  • Registered Users Posts: 65,162 ✭✭✭✭unkel
    Chauffe, Marcel, chauffe!


    Originally posted by Sir Random
    I wouldn't run it 24/7 at 1.7v on air, P4s have suffered sudden death from overvolting

    Not at all :)

    You're grand running at 1.7V - P4C won't die, just slow down when overheated

    Ok, the cpu might not last 20 years, but how relevant is that?

    Go for it m8 - happy overclocking :D

    P.S. amazed at the overclock you're getting already at stock voltage :eek:


  • Registered Users Posts: 968 ✭✭✭Adeptus Titanicus


    K, had it at 1.7v, 247FSB running Prime last night, but it only ran for about an hour before the system locked up. As I mentioned before, the temps look pretty ok, I was just wondering would something else get zapped. Might do like Random did and pop it right up for the hell of it, just to get some benchmarks at that speed. (was your 4GHz stable at that voltage Random?) Actually, I haven't benchmarked it at all at 3.7GHz. Might do that before running Prime again.

    I'll try a slightly lower FSB at 1.7 and see how it goes. Is it worth risking upping the vCore a little more? :)

    And yep! I'm fairly pleased at my 3.6GHz at stock voltages! I shouldn't be so greedy!:)


  • Registered Users Posts: 65,162 ✭✭✭✭unkel
    Chauffe, Marcel, chauffe!


    Originally posted by Adeptus Titanicus
    I'll try a slightly lower FSB at 1.7 and see how it goes. Is it worth risking upping the vCore a little more? :)

    Sure just for the crack. Like I said, you're not gambling with the life of your cpu
    Originally posted by Adeptus Titanicus
    I shouldn't be so greedy!:)

    It's only natural ;)


  • Advertisement
  • Registered Users Posts: 771 ✭✭✭Sir Random


    Originally posted by unkel

    quote:
    Originally posted by Adeptus Titanicus
    Is it worth risking upping the vCore a little more?

    Sure just for the crack. Like I said, you're not gambling with the life of your cpu
    How can you advise someone to go over 1.7v so flippantly? Never heard of SNDS? He will be gambling with the life of his cpu if he runs it over 1.7v 24/7 on air. If not, we'd all be doing it.


  • Registered Users Posts: 771 ✭✭✭Sir Random


    Originally posted by Adeptus Titanicus
    Might do like Random did and pop it right up for the hell of it, just to get some benchmarks at that speed. (was your 4GHz stable at that voltage Random?)
    I never set my vcore above 1.65v in the BIOS, and 4Ghz wasn't stable (just did minor benches).


  • Registered Users Posts: 968 ✭✭✭Adeptus Titanicus


    Originally posted by Sir Random
    I never set my vcore above 1.65v in the BIOS, and 4Ghz wasn't stable (just did minor benches).
    k, thanks. Like I said, I never benched it at 3.7, so I might just do that. TBH I wouldn't feel too safe about running it 24/7 on 1.7, so it's just for the sake of seeing how high I could potentially go. Do you think the main issue would be heat? i.e., if I had water colling would it be safer? Or do you think it is it just a case that running it this high is just not good for it?

    Thanks for the opinions guys. But be assured, I won't be rushing into popping my vCore up to 2 or something.. :)

    PS, what's SNDS? I'm pretty sure it's not the Sisters of Notre Dame, or Special Needs Design Services.. :)


  • Registered Users Posts: 65,162 ✭✭✭✭unkel
    Chauffe, Marcel, chauffe!


    Originally posted by Sir Random
    How can you advise someone to go over 1.7v so flippantly? Never heard of SNDS? He will be gambling with the life of his cpu if he runs it over 1.7v 24/7 on air. If not, we'd all be doing it.

    Obviously I'm not advising to just up the voltage from 1.5 to 1.8 volt, not check the temperatures and then leave the house for the weekend :dunno:

    Upping the vcore in small steps while keeping a good eye on the temps on a P4C with good air cooling and then leaving it @1.7V provided all is well under load will not lead to cpu failure. SNDS is very rare anyway on P4C unlike it's predecessor and even more so at just 1.7V


  • Registered Users Posts: 968 ✭✭✭Adeptus Titanicus


    Well, I ran a load of benchmarks on it this evening t 3.75GHz.

    Tried 250 FSB on 1.7v with the RAM at 5:4 2.5-3-3-7 @ 2.65v
    then 250 FSB on 1.7v with the RAM at 1:1 at 3-4-4-8 @ 2.65v as it didn't seem to want to boot at tighter timings. Might try it again later.

    The thing is it ran very well for 2 each of 3D01, 3D03 and Aqua03 (benched each twice, one without the gfx o/c (9800XT stock levels) and one with an o/c(327.5/405)). I'll try prime again with these settings. Timings can make a big difference with some of these benches...

    Anyway, just in case anyone's interested :)

    Stock everything for reference
    200 FSB RAM @ 1:1 2.5-3-3-7, gfx @ 9800XT stock
    3DMark01: 17,691
    3DMark03: 6,439
    Aqua03: 44,997
    SuperPi: 47secs

    My normal OC
    240 FSB, RAM @ 1:1 2.5-3-3-7, gfx @ 9800XT stock
    3DMark01: 19,911
    3DMark03: 6,612
    Aqua03: 48,108
    SuperPi: 40secs

    250 FSB RAM @ 5:4 2.5-3-3-7, gfx @ 9800XT stock
    3DMark01: 19,987
    3DMark03: 6,610
    Aqua03: 47,839
    SuperPi: 39secs

    250 FSB RAM @ 1:1 3-4-4-8, gfx @ 9800XT stock
    3DMark01: 20,234
    3DMark03: 6,632
    Aqua03: 48,136
    SuperPi: 39secs

    So, at a 5:4 ratio, going up to 250FSB doesn't give me much of a boost. Likewise, 1:1 at 250 with looser timings don't really amount to alot either, just a couple of hundred marks on Aqua03 and 3D01, but still a little better than 5:4. I'll try with tighter timings again after the weekend (taking a well-earned break ;)). I'd say tighter timings will make a bit of a difference. Any suggestions bearing in mind it's the 3700 Rev2?


    Oh, and here's the 250 with the gfx upped a bit as a matter of interest
    250 FSB RAM @ 1:1 3-4-4-8, gfx @ 327.5/405
    3DMark01: 20,716
    3DMark03: 6,991
    Aqua03: 50,029


  • Advertisement
  • Registered Users Posts: 65,162 ✭✭✭✭unkel
    Chauffe, Marcel, chauffe!


    Adeptus Titanicus, interesting results :)

    What was vcore during the testing? It looks like your normal overclock is pretty much optimal. With the P4C and high FSB settings, most people (including myself) would report that running the ram at 1:1 is best if the timings are ok for high speed memory. Most benchmarks online only compare low speed, tight BH5 2,2,2,5 to high speed loose 3,4,4,8 and don't report as what seems to be best for P4: high speed, tight enough 2.5 memory

    Your ram is obviously holding you back a little although I'm amazed it runs as PC4000 with loosened timings. If you had my PC4200 OCZ, you'd surely do 250FSB or more with 2.5,3,3,7 or 2.5,4,4,7. How far would that let you go?

    Your mobo/cpu combo is working better than mine. I can run XP @250 and can run some benchmarks, but no prime 95 stability :(


  • Registered Users Posts: 968 ✭✭✭Adeptus Titanicus


    Originally posted by unkel
    What was vcore during the testing?
    ...

    Your ram is obviously holding you back a little although I'm amazed it runs as PC4000 with loosened timings. If you had my PC4200 OCZ, you'd surely do 250FSB or more with 2.5,3,3,7 or 2.5,4,4,7. How far would that let you go?

    Your mobo/cpu combo is working better than mine. I can run XP @250 and can run some benchmarks, but no prime 95 stability :(

    Ah, but I never said it was prime stable ;) Actually, I don;t know if it is or not at these settings. When I first started overclocking this I tried 250, but it wasn't prime stable, but I hadn't messed with the timings, so I'll have to run that again. Like I said though, it ran 6 benchmarks no problem, while when I had it at tighter timings the first test I ran, Aqua03, froze up.

    vCore at 250 was 1.7v, otherwise 1.55v. Temps during benching never went above 51.

    I might try slightly tighter timings too. 2.5-4-4-7 looks interesting. I know from experimenting that going from 2.5-4-4-8 to 2.5-3-3-7 at stock speeds makes a fair difference.

    Ta

    AT


Advertisement