new true430, giving it another go
alright so now that i've got my new true430, and i can actually read my temp sensors correctly (yay new MBM version) its time to give OC'ing this 2.8 another go. so far i have had absolutely 0 luck, however that may have something to do with the fact that i cant get it to give more than 1.48v to the CPU... stupid cheapass allied 450w.
anyway i think the corsair 3500 is good, especially with the dividers in use, i know the IS7 ought to be good at least for up to 255, and i'm HOPING this 2.8 isn't a pos chip.
i'll keep you guys updated, hopefully my slk-900 will finally have to earn its keep
anyway i think the corsair 3500 is good, especially with the dividers in use, i know the IS7 ought to be good at least for up to 255, and i'm HOPING this 2.8 isn't a pos chip.
i'll keep you guys updated, hopefully my slk-900 will finally have to earn its keep
0
Comments
fsb to 216
mem 1:1 @ 2-3-3-7
vcore 1.6
vdimm 2.65
post, but no windows
dropping vcore to 1.575 = windows, but immediately get an IRQL_NOT_LESS_THAN_OR_EQUAL
set mem divider to 5:4, mem @ 174
worked
odd
dropped gat from turbo to auto
raised mem back to 1:1
got a siren
wtf?
VDIMM @ Default.
CPU @ Default Volts.
Crank your FSB up to 216 and Prime95 test. If it passes, continue cranking the FSB until you find the max of your chip.
If you are on air cooling with the P4 (high end, like the SLK900 and a good 92mm fan), you can safely crank the VCore to 1.625V, but I wouldn't go over that. The IS7, IC7 & IC7-MAX3 all undervolt like crazy and need that extra bit of juice.
If you want to run your RAM @ 1:1, set to default voltage, drop timings to 3-4-4-8 and start from there, slowly tightening timings and increasing voltage to your safe limit.
its 2x256 corsair xms3500, rated at 2-3-3-7 @ ddr433 (216 right?)
so the question is WHY do i need a memory divisor at 216, it ought to work at 1:1
on the plus side, apparently the wierd high pitched noise my comp was making was coming from the PSU ( i couldn't hear it anyway
UPDATE:
220 seems to be good to go
thanks a bunch simguy, i was under the impression i needed to raise vcore in order to find the chips limit, then back it down from there. this seems to be working better (although its still undervolting like crazy [it IS .02 v better, which is nice], at least i know i dont have a weakass chip)
UPDATE 2:
seems my chip hits a wall at 235, which as far as i know isn't too bad for a 2.8. still, some of those 2.4's can go farther, sucks
3311A278 (may have a -XXXX on the end)
What do you have VCore & VDIMM set to for that 235? I would think you should be able to go higher than that...
edit :
the chip posts all the way up to 250 (didn't try higher) , and will boot into linux just fine with whatever vcore i set it at. however, the little *******refuses to boot into windows with a vcore higher than 1.575, regardless of the FSB.
on top of that, despite my ram SAYING its pc3500 and being corsair, i can't get it to boot 1:1 at even 216. i can set the timings to 3-4-4-8, set GAT to auto, and even bump the vdimm to 2.8, and nothin is doin.
i checked the voltages (with both PSU's). vdimm is consistently .1v higher than it should be, and vcore is anywhere from .025 to .075 under depending on what voltage i set.
it looks like the reason i haven't been able to run this corsair 1:1 is because of abit's GAT. i had been trying to get the ram to work on "Turbo" which is just basic enabled GAT, which i would think would be just the same as if you were using an actual PAT enabled board and left everything at default. it would seem this is not the case.
as such, everything works fine and peachy prime stable at 1:1 up to 220.
at 230, i no longer get a memory siren, but instead windows gives me some message about a corrupted system file... and tells me to do a repair install. wtf IS this? any tips appreciated