SurClaro Flight Simulator Resources

Posted: Tue Feb 26, 2008 11:50 pm Post subject:

Wait...did you just say any voltage under 1.524V will cause BSOD?

Remember, an increase in voltage will heat up the processor rather quickly. I would always increase the FSB.

Posted: Wed Feb 27, 2008 9:08 pm Post subject:

I'm saying it takes 1.52V to run at 3.51GHz. Is it safe to keep my voltage at this level? Is it safe to keep my voltage at 1.520V? Intel stated the highest voltage recommended is 0.89V to 1.5V. So I'm just a little concerned, lol.

Captain El Razz

Posted: Thu Feb 28, 2008 2:25 am Post subject:

Your current voltage is at 1.3V. Try 1.4 and increase the FSB to 353 MHz (this was your initial desire, stated way back in the beginning of this topic). This will yield 353*9 = 3177 potential MHz. Tell us what happens.

When overclocking, always increase any number by a very small factor. Then do a stress test. This will guarantee stability. While your processor may tolerate 1.5V and a clock speed of 3.51GHz, you are pushing the limit and may risk a sudden lock-up. Obviously, you don't want to be playing a game or editing an HD movie then have the processor fail on you.

EDIT: I suggest you look up this product on Newegg.com and check out the user reviews. I bet there will be a person who achieved a massive overclock.

Posted: Thu Feb 28, 2008 8:18 am Post subject:

I did that. I am now at 9x 390MHz -> 3510MHz (3.51GHz), but I require a voltage of 1.520V in order not to get a BSoD after running Prime 95 with small FFTs for an hour or so. I'm running my Q6600 at 3.51GHz rock-stable at this 1.52V. But my question is... is it safe to run my CPU at this voltage in terms of everyday usage?

Captain El

Posted: Thu Feb 28, 2008 7:28 pm Post subject:

Keep running stress tests, maybe once a day for an entire week. Then you can slow down that rate. This ensures stability for a long period of time.

This is interesting stuff, El. You said you had to increase the voltage to 1.52V in order to achieve 390MHz FSB. How did you figure out the optimum voltage?

Posted: Thu Feb 28, 2008 9:45 pm Post subject:

I ran a test in Prime for 16hours 37minutes hours last night, no problems Very Happy

.

It's perfectly stable at 3.51GHz at 1.520V. I got this optimum by raising the voltage one tiny step each time I got a BSoD while running Prime. It's a painstaking process... but I now have the optimum Smile

.

But I am still wondering if 1.52V is a safe margin to run a Q6600? I do not want to slowly damage my processor due to slight overvoltage Confused

Captain El

Posted: Thu Feb 28, 2008 11:28 pm Post subject:

OK. Correct me if I am wrong. I am going to decipher the process:
1. Increase FSB to desired value
2. Bump up voltage slightly.
3. Do stress test
4. If BSOD occurs, go back to Step 2. Repeat process until system is stable.

Any type of CPU overclocking is pushing the limits, so you are decreasing the life of the processor. Eventually, the processor's gonna die. It also depends on the processor you received (some receive defective processors, while some receive highly overclockable and robust processors). We can't make any final judgments. Just know that your task will decrease the life of the processor.

Hope that all makes sense.

Posted: Fri Feb 29, 2008 8:58 am Post subject:

Yes. Except I would add another step to your method Wink

.

I'd start from my original clock and start with +5 FSB increments increase each time. Keep your voltage where it is. Run Prime for 16 minutes or so (small FFT's!) is sufficient to simulate for most daily tasks (unless your running superheavy games). And keep adding +5 MHz until you get an error in Prime before 16 minutes or BSoD before 16 minutes. Then it's time to raise the voltage one increment. To make it simple:

1. Increase FSB +5 MHz increments from your original speed
2. Stress Test 16mins small FFT's in Prime
3. If no problems occur, go back to Step 1. If BSoD or problem occur, continue to step 4.
4. Bump up voltage slightly.
5. If no problems occur, go back to Step 1 (or step 6).
6. If at desired speed and no problems after 16mins. Continue Prime test small FFT's for 3 hours.
7. If no problems occur, continue to run for Prime for 10 hours (you might want to leave it on overnight)
8. If BSoD, bump up voltage slightly, and repeat for 10 hours.
If no problems after 10 hours, continue run 16 hours.
9. If BSoD, go back to step 8.
10. If no problems after 16 hours. CONGRATULATIONS! You have a stable overclocked computer Smile

It is a painstaking process, take my word for it.. lol, it took me about a week to find out what my stable speed should be. But the results are worth it Wink

. I run my Q6600 at 3.51GHz without problems. I am aware that it decreases CPU lifetime, but I am just wondering if enduring operations above 1.520V will cause premature burnout?

Captain El

Posted: Fri Feb 29, 2008 9:33 am Post subject:

El,
Thanks so much for the process! It all makes sense now!

"Premature" is a bit broad. It could be tomorrow, or it could be 15 years. We will never know the expected life of a single processor, thus it is futile to think about when the processsor will die. FYI, generally, the processor WILL burnout prematurely, only because you have overclocked it. Just my $.02

Posted: Thu Mar 06, 2008 8:00 am Post subject:

I know that overclocking will decrease CPU lifetime overall just for doing it. But I am still wondering if there will be any particular danger in overvoltaging the CPU above 1.52V?

Captain El

Posted: Thu Mar 06, 2008 9:47 am Post subject:

As long as it is stable at 1.52v, for lets say 2 hours of working it hard, and no BSOD's, then you should be OK.

Just remember, it will take away time. Instead of being good for 3 years, it could be good for only 3 weeks, or 3 months.

RD

Posted: Wed Mar 12, 2008 8:23 am Post subject:

3 weeks to 3 months? Are you sure?

I ran mine already for a long time... and no problems.

Posted: Wed Mar 12, 2008 8:37 am Post subject:

It really depends on alot of factors, and how good the components are.

So yeah, it could last 3 weeks or 3 months. If your lucky to get a good one, it could last 3 years. There's alot involved, that you and I can't see. And that is stability and quality.

RD

Posted: Thu Mar 13, 2008 3:48 am Post subject:

Posted: Fri Mar 14, 2008 7:02 pm Post subject:

Hmm, I agree. But maybe 3 weeks or 3 months is a bit exaggeratedly short. Maybe 1.5 to 3 years seems more reasonable, because if my CPU is in use, it'll run at 3.51GHz.. but when its idle, it'll 'throttle' down to 2.3GHz (from 9x multiplier to 6x, as of why I thought mine was underclocked to 1.6GHz in the beginning Laughing

).

So what'll happen to my CPU if I ran it at 4.01GHz at 1.76V?

Captain El Razz