IS7 Pat-Enabler forces 5:4 Divider?

TheBaronTheBaron Austin, TX
edited December 2003 in Hardware
I've recently started fooling around with some OC'ing, despite poor results in the past. I was under the impression that my chip was weak as was my ram, but some sneaky results ( :hiding: ) have shown that it might just be that this board sucks ass. here we go...

IS7 ; 2.8C (sorry, no week #'s) ; 2x256 pc-3500 corsair ; antec tru430 ; slk-900 ; enermax adj 92mm fan

question #1 : according to CPU-Z, when the IS7's pat enabler is ON, the board forces a 5:4 divider no matter what i have the actual divider set to (SPD, 1:1, 3:2). this is aggravating, to say the least. is it possible CPU-Z is wrong, or is something seriously messed here?

question #2 : as far as i can tell, the system will only finish booting into windows at 3 vcore settings - 1.525, 1.550, 1.625 ... nothing else. any ideas?

Comments

  • ketoketo Occupied. Or is it preoccupied? Icrontian
    edited December 2003
    #1 definitely shouldnt be happening that way. Are you on the latest bios?

    #2 same story, that's a very strange one
  • TheBaronTheBaron Austin, TX
    edited December 2003
    i'm running bios 17, which i flashed using the windows utility

    here's what i'm thinking
    either the board is messed, or the memory is messed, and here's why: the memory registers as pc3200 (again cpuz) and always gives errors at the advertised timings at 216 (which = 433 = pc3500 riiiiiight?). however, i was planning on buying some more ram anyway, and the hyperX looks like a good deal, so pretty soon i can rule out the memory as the culprit.
    i definitely dont like this 5:4 divider business, so at the moment i have it set to "auto", no pat enabler.

    onto the subject of the voltages, i figure this is also either the CPU or the board. the cpu is most likely very weak, because i cant even do 215 and still be prime stable, even at 1.55v (one notch up). 210 seems to be okay at 1.55v, but this is not the kind of overclock i was hoping for (i was thinking more along the lines of 250)

    however, on principle i refuse to run a 5:4 divider at anything under 250 (i want at least ddr400 speeds damnit), especially since not doing so eliminates the performance increase the overclock was supposed to give anyway. therefore, even though i can get the chip prime stable at 230 @ 1.625v with a 5:4 divider, what the hell is the point if my memory is running at ... less (now is not math time)

    stupid ass crap
  • ketoketo Occupied. Or is it preoccupied? Icrontian
    edited December 2003
    Actually, if you bench your system 230fsb 5:4 PAT enabled, I think you'll find it substantially faster than 210-215 1:1 no PAT. This doesn't solve the problem, of course....hmmmm. That chip *should* do 230 all day long at 1:1 and the corsair *should* do it - my PC3500 Corsair will do 230 fsb, albeit at very relaxed timings on my P4C800-E Deluxe w 2.6c. Hell, your chip probably IS capable of 250 fsb, or close. Have you tried changing RAM slots?
  • TheBaronTheBaron Austin, TX
    edited December 2003
    nah, i'll give that a shot though, i guess you never know

    UPDATE:
    turns out setting GAT from Turbo to Auto (turning PAT off) seems to fix my silly voltage problem. it also seems that when a 5:4 divider is manually set, PAT doesn't get enabled at all, no matter what GAT setting i choose. i think this board is weak sauce, i want a MAX3
  • ketoketo Occupied. Or is it preoccupied? Icrontian
    edited December 2003
    TheBaron wrote:
    UPDATE:
    It also seems that when a 5:4 divider is manually set, PAT doesn't get enabled at all, no matter what GAT setting i choose. i think this board is weak sauce, i want a MAX3

    That sounds more like what I had read on other forums. I still think you have *hopefully fixable* RAM problems. Now that you've figured out 5:4, time to go back and try 1:1 again see if anything you've changed adds new stability?
Sign In or Register to comment.