not enough... OCCT test? 10 loops IBT arent enough....
run your windows in diagnostic mode to free up more ram...then run IBT for 1~2 hour
also run OCCT test for 1 hour... 3dmark few loops...
trust me...passing prime95 for few hours doesnt make your PC 100% stable...

This post has been edited by campsol2k: Jun 5 2009, 08:52 PM

i know. but as long i didnt get any blue screen that consider stable for me.

ok.. BSOD is just worst case senario....but don't complaint if your games or apps crashing in windows...

This post has been edited by campsol2k: Jun 5 2009, 09:21 PM

so far gaming no problem. if get the bsod just clock down then the prob settle edi.

if i clock using the ways i usually use then i have no prob... i can pass IBT 10loops.. and i always test using IBT, lazy to wait for prime95. lol...

but after hilmiangah posted that way of stability test.. i found out that my OC is not 100% stable. IBT cannot be completely trusted. 8mins = 40hours of prime95. lol... Quite true in fact, but just not PERFECT.

But the tips telling u what to add when errors happened from the OC stability test that hilmiangah posted, is hard to determine which setting is the culprit. So instead of using that. i used the way from this link

http://www.tomshardware.co.uk/forum/page-240001_11_0.html

I wrote a log of my settings. Doing different iterations.

Im doing experimenting it. Doing quite well.. still in progress to find the lowest settings possible.

why people always get the wrong idea that 8 mins IBT equals to 40 hours Prime95...? 4me its not equal at all...
it doesnt matter u ran prime95 for 40 hours..
u must complete the OCCT test at least 1 hour for each test set(small,medium n large) and IBT test (max ram) at least 1 hour... including other tests(prime95, etc) of course...
thats the common rule....
B4 IBT was released to public/overclockers community... i use OCCT as my main tester instead of prime95.....
believe it or not i ran 7 hours prime95 but failed OCCT test within 15 minutes...
the test methodolgy made by cstkl is just to pinpoint which voltages are needed be increased...thus reduce time and effort... to stabilize your overclocked setting... without having too many tester...
i think its a good guide.. but i'm not sure about the accuracy of the relationship between each voltage and each custom prime95 torture test...
neither pictures nor overclocking logs are provided by cstkl in the guide..

This post has been edited by campsol2k: Jun 6 2009, 06:50 AM

dude... ppl are gettting the wrong idea is because it's written in the readme file. LOL..

That's why i said. IBT is not perfect... getting pass 10loops or 20loops is just some sort of placebo effect that ur pc is super stable which is not... IBT is very demanding and also DAMAGING... i will not run IBT for one hour... I will only use IBT for earlier stages of stability test.

For me, getting IBT 8 ~10 loops is not even CLOSE to ROCK HARD STABLE.

Read the link i just posted, the time of getting settings for ROCK hard stable OC needs is reduced a lot by using a log instead of guessing the values of what ur motherboard needs.

Just a simple comparison between boards.

Jetway p45 xblue 7 x 400 everything on default fails prime in 1 hour. - blend test

Gigabyte EP45UD3L 7 x 400 everything on default passes prime for 20 hours,manually stop - blend test (doesnt mean stable but at least better than the other board)

how good a mobo is, really affects a stable OC FOR REGULAR use.

I can anytime.. get 3.8ghz~4.0ghz using xblue p45 for benchmarks such as 3dmark06 or 1~32m super pi. but getting a stable OC is really hard on the xblue p45 for me. As for gigabyte EP45UD3L, it's really easier...

I do agree with u...
i've experienced myself... clock 3.9GHz Q9550... multi 8.5 vtt 1.28, vcore 1.275, vnb 1.26 vmem 2.1
for example... i've
-passed 38 run LinX(approx 10 minutes) then crashed
-passed Prime95-cstkl Test 1 and Test 2 but crashed in Test 3
specifically for FFT size 448K.. ran custom test again specific for FFT 448.. then crashed b4 1st interations was completed...
-passed OCCT small set for 1 hour...

So 8 mins IBT isnt equal to 40 hours Prime95... its a flawed statement...

This post has been edited by campsol2k: Jun 6 2009, 07:06 PM

i agree with u. ibt really stress the processor. i'm afraid it will damage the processor if running for 1 hour or more. even with adequate cooling.

That's an excuse not to run LinPack for 1 hour and nonsense thinking... if u kept everything cool below recommended limits ie VRMs, mobo, CPU, NB, SB... ur system won't die easily ..
the effect of increasing your voltage too high and too much exceed recommended/max voltages that will permanently damage your hardware because of electromigration... even with adequete cooling...

This post has been edited by campsol2k: Jun 7 2009, 07:26 PM

er... seriously lor.. IT's not a MUST thing ok... most of the ppl just run OCCT and prime out there.. there's no MUST this and MUST that thing.. it's only self preferences...

just a quick test

wah low temp..

tkh, did u fap to ur low temp?

lawl... i came many many times when i see the voltage needed for that clock

???? came as in that "came"? hahaha... btw not bad wor ur temp at such vcore... TRUE performs that well ker? @@ mine 1.17v @ 3.6ghz oso oredi same temp as u, using ifx-14 with 3 fan summore @@

lawl.... i dont know man.... maybe i am on naked rig punya pasal kot

Lol. I still remember tkh bro's old q9450 c1. Needed 1.4v to reach 3.6ghz right?

No wonder u came this time la lol.

'I cummed many many times when i see the voltage needed for that cock'

wait wait... dont come 1st......

i still trying 4ghz... but then, no tweak applied to gtl since too busy nowadays....

just simply whack vcore, vnb, and vfsb lol

i just realize ur minimum core temp is 23c... lol... u use aircon blow ah? hahaha..

could be he take dry ice then put fan in front of it hahaha