q9550 at 3.4 want 3.6, quick question.

[alwaysoc0nfused]

Hey guys, as some of you might know from my previous posts i have the following now:

-EVGA 780i
-q9550
-tx750 psu
______________

So after a few weeks of trying and returning my 750i board and switching for another q9550 I achieved a overclock of 3.4 stable. Settings are as follows;

CPU core: 1.275 (1.24 in cpuz)
CPU FSB: 1.3
Mem: 1.85

I passsd 1hr OCCT, 40 runs of max Intel Burn Test, and over 8hours of prime95. Air cooled and temps are in the high 20s at idle and on heavy load(OCCT, IBT and Prime95) it maxed out at 57c. I feel like I should be able to get 3.6ghz safely but every time i increase the FSB it fails immediately.

I have tried to increase the cpu core volts a bit but nothing. Should I up the FSB voltage a bit? I have read and researched a million times and i feel like since i have tried upping the core volts a bit with no success i should try uping the FSB volts a bit. I just want your opinion before moving forward as i do not want to screw things up. Thanks.

 

[alwaysoc0nfused]

bump, any help at all?

 

[kyeana]

How high did you increase the vcore to before? Probably it just wasn't high enough.

 

[alwaysoc0nfused]

i went 2 notches at a time to all the way to 1.35 and even then it BOSD 10 seconds into intel burn test.

 

[Craxbax]

Try your vcore to 1.4v and see if that works. What you set is bios will be lower due to vdrop and vdroop at load. I am not familiar with your boards bios but you may need to play with GTEL REF/PLL/VTT voltages. You may also need bump up your NB voltage especially if you have 4 gb of ram.

 

[alwaysoc0nfused]

Ok should I:

a: leave cpu volt at 1.275 and up NB volts first

or

b:up cpu volts to 1.4 , with vdroop it will be .3 less in cpuid.

 

[mtyermom]

Why the obsession with getting to 3.6 over 3.4? I mean, I understand trying to get every last bit of performance. But I mean, 3.4 with a max load temp of 57 and completely stress test stable... ain't so bad. It's possible that you've hit the wall for your particular proc.

Keep trying though, I'm curious to see if you'll find the magic combination. Though, I'm wondering if you'll be able to keep it within comfortable thermal levels on air cooling as you keep increasing that voltage.

As to your last question, I'd say set vcore to 1.4. Then gradually bring up NB voltage until it's (hopefully) stable. Once there you can gradually bring the vcore back down (if at all).

 

[cd14]

The CPU will need more FSB voltage (stay below 1.50) and vcore. I'm at 4.08 with mine on a P5Q. My memory test fail if I try to go more. It's still within spec so I'm not sure if the board is reaching FSB limits.

 

[silko79]

I have mine stable @ 430X8.5 = 3656MHz (OCCT, Prime95, BurnIn, 3DMark06)
FSB=1721MHz
Vcore 1.4Volt (set in BIOS) gives 3.6Volt in CPUD
DRAM 2.2Volt
Temperatures:
Idle 36C MAX 57C

 

[alwaysoc0nfused]

okay after a lot of messing around here is what i got;

stable 3.6 fsb at 1700, i had it stable at 1750 but i figured there is no difference
vcore: 1.368 (shows 1.32 on cpuz)
vtt: 1.3
temp:
idle 28c max 65c(on the highest core)

I feel like i have reached my limit, i can bump the vcore up but i cant get it stable at anything higher than 1750 fsb.

am i safe running at 3.6ghz 24/7?

and since i am not stable anything higher than 1750mhz and i only tried going as far as 1.4v should i try to bump vtt up to 1.4 to see if i can get stable?

 

[kyeana]

you could bump up more settings such as vcore, vtt, north and south bridge to try to get it stable higher, but unless you just want to see how hard you can stably push your processor you will see almost no gains from the extra overclock. If your happy with how the system performs and aren't looking to find the absolute maximum overclock you can achieve, 3.6 should suite you fine

 

[alwaysoc0nfused]

Achieved 3.6 stable.