• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Need help OCing

Gensou

Limp Gawd
Joined
Dec 8, 2007
Messages
344
Hey [h] fellas.

First, what I'm running:

q6600 g0 1.3125v vid w/ zalman 9700led
evga 680i sli a1 mobo
4gb ddr2 800 ( corsair xms2 )
750w psu thermaltake
2x 8600gts sli @ 700/1670/1010 w/ zalman vf700

ok, i've been running at 3.2ghz all this time and i was reading through some posts and such and i wanted to try to get this baby to at least 3.6ghz.. think i can do this?
I think i have a huge vdroop.
right now i'm running at 3.3ghz ( 366.67 x 9 ) w/ 1.352v at full load. my temps are 54c load.

all my voltages are auto except vcore which is set to 1.42500 ( i think ) in the bios.

i was wondering if anyone can take me step by step to achieving 3.6 stable
i have no clue what mcp voltages are and what to do with them and i need to increase them.

if you are willing to take your time and help me i will be very appreciative.
if you need screenshots and more detailed information just say so. I will try to provide all the information I can.

thanks.

btw, sorry for this wall of text. :p
 
Hey [h] fellas.

First, what I'm running:

q6600 g0 1.3125v vid w/ zalman 9700led
evga 680i sli a1 mobo
4gb ddr2 800 ( corsair xms2 )
750w psu thermaltake
2x 8600gts sli @ 700/1670/1010 w/ zalman vf700

ok, i've been running at 3.2ghz all this time and i was reading through some posts and such and i wanted to try to get this baby to at least 3.6ghz.. think i can do this?
I think i have a huge vdroop.
right now i'm running at 3.3ghz ( 366.67 x 9 ) w/ 1.352v at full load. my temps are 54c load.

all my voltages are auto except vcore which is set to 1.42500 ( i think ) in the bios.

i was wondering if anyone can take me step by step to achieving 3.6 stable
i have no clue what mcp voltages are and what to do with them and i need to increase them.

if you are willing to take your time and help me i will be very appreciative.
if you need screenshots and more detailed information just say so. I will try to provide all the information I can.

thanks.

btw, sorry for this wall of text. :p

400*9=3600 With a 1:1 ratio it will keep your ram at stock speeds. Set your voltage to 1.525V and start running intelburntest, If its stable for 20 runs, start decreasing voltage until its not 20 runs stable and you got the voltage you need.

Im looking up your bios and i will give you some detailed inputs to put in the bios and try out. Hang on for like 5 mins
 
According to bittech your bios is very straight forward, so its just the matter of setting a fsb of 400 and multi of 9, and setting the voltage.
 
20 runs? How long is that? I've never used intelburntest. Also, 1.525? Isn't that a bit too much.

I just did the pencil mod and now my voltages are like this:

in Bios: 1.416
load: 1.416
idle: 1.408

Is that good?
 
20 runs takes about 25-30 minutes.. not very long..

1.52v is a good starting point on 65nm its not a problem.. basicly its high enough to know the overclocks stable.. then you work your way down til you find the lowest possible stable voltage..
 
20 runs? How long is that? I've never used intelburntest. Also, 1.525? Isn't that a bit too much.

I just did the pencil mod and now my voltages are like this:

in Bios: 1.416
load: 1.416
idle: 1.408

Is that good?

Depends on what you running, if your on 64 bit on full stress, Its about 6-8 mins per run, So make it 5 runs since your testing stability and whats the voltage needed for 3.6ghz, Once it does not pass 5 passes @ 3.6ghz on the voltage your testing, turn it back to the previous voltgage setting and then run 20 runs. Are you on 64 bit vista/xp? If so, Just open it, Choose no for error detection, 1 for full stress and then 5 for 5 passes. Make sure the Residual <norm> come all out the same, if even 1 number comes out different its not stable.

1.525 is not much for 65nm, The max safe voltage for a 65nm is 1.55V, Some say/feel its 1.5V. Remember i said to start at that voltage and just keep turning it down til its not stable.

Yes the voltage is very good if its 3.6ghz stable, Test to see if its 20 runs ibt stable. For me, full stress uses about 3.6gb of ram so it stresses ram very good as well, but then again, if you are running the settings i said, ram should be a stock and not a problem.

20 runs takes about 25-30 minutes.. not very long..

1.52v is a good starting point on 65nm its not a problem.. basicly its high enough to know the overclocks stable.. then you work your way down til you find the lowest possible stable voltage..

Damm wth you running like i said mine takes about 6-8 mins per run and uses a shit ton of ram, makes me feel very confident its stable. :cool:
 
I hope intelburntest doesn't kill my chip. I'm already getting 54c in prime95.

Also, which stress level do i do?
1 - maximum stress
2 - half stress
3 - low stress
4 - customize stress level

Btw, I'm on xp 32bit.
 
1. Maximum stress.

If you are getting 54c with P95 you can probably expect temps in the upper 60's to low 70's with IBT, which is perfectly safe. It's pretty common to see temps 10-15c higher with IBT than P95. The chip is safe to 100c where it will start to throttle the multiplier and voltage. I wouldn't worry about killing the chip.

The only thing you may run into issues with is getting stable at 3.6 with a 680i board. I went through 3 680i boards before I gave up and grabbed a P35 based board and had much better luck.
 
Ok, I'm running it right now @ 3.3ghz with 1.41250v. Highest it has hit so far is 61c. It's on 4th trial right now.

----------------------------------------------------
Testing finished in 410.078 seconds.

All Residual<norm> values are the same.

So if I do 1 test with 5 trial runs and I pass...does that mean I'm pretty stable?

Just did 20 trial runs in 1792 seconds ( 29mins ). Am I stable?
 
Check that all the residual <norm> numbers are all the same, if they are, your stable. If even 1 number is off, its not stable.
 
Back
Top