View previous topic :: View next topic |
Author |
Message |
eamo

Joined: 08 Mar 2003 Location: Shepherd's Bush, 1964.
|
Posted: Wed Nov 14, 2007 4:40 am Post subject: Okay. *gump* Overclocking question....... |
|
|
My new and awesome PC is in the pipeline so I thought I would mess around with my old system and try my hand at overclocking for the first time..........
................Don't worry! Nothing has blown up and all seems well....so far.
The thing is, I can get the CPU to the high frequency speeds boasted of all over the tech forums on the net, but not at the low/nominal voltage those guys claim. Extra voltage causes extra heat and wear and tear. Right?
I've simply been raising the FSB multiplier.
Spec check...
Intel 930D (3.0ghz) (nominal voltage range...1.25-1.4V)
Gigabyte GA-945P-S3 (rev. 3.3) (667 FSB)
Samsung PC2 5300 (667) 2X1gb (dual channel) (running at 640mhz. FSB:RAM ratio 3:4. I slightly underclocked the RAM to get better timings and it has worked according to benchmarks).
I've got the CPU to 3.6ghz with very agreeable temps. Idle 34. Never over 58 on 100% full load. The Zalman CNPS9500 helps a lot there.
The problem is that the voltage on the CPU has risen to 1.48V. Over the spec and a lot higher than the guys on the net claim to need. Some credible forum members say they got this chip to 4.0ghz on the nominal voltage of 1.2V.
How can I overclock without raising the voltage so high?
I'm aware that maybe less than 5 people who browse this forum might know something but it's worth a shot. |
|
Back to top |
|
 |
kprrok
Joined: 06 Apr 2004 Location: KC
|
Posted: Wed Nov 14, 2007 6:08 am Post subject: |
|
|
I don't know a thing about OCing because I've never tried it. I figure I don't do anything taxing enough to worry about it.
That said, I'd be interested to hear how to do this should I ever decide it'd be worth it.
KPRROK |
|
Back to top |
|
 |
SeoulFinn

Joined: 27 Feb 2006 Location: 1h from Seoul
|
Posted: Wed Nov 14, 2007 11:55 am Post subject: Re: Okay. *gump* Overclocking question....... |
|
|
Is 3.6 your max? 600MHz for free is quite good. CPUs are unique. The highest OC varies greatly within the same batch (stepping). Some go faster, some don't. Some need more juice than others. Maybe your CPU is one of those that need more juice to go even that far?
I'm not sure how you could lower your voltage. These things tend to run faster the cooler you get them. Since your temps are already pretty good, maybe those other guys used water cooling in their rigs?
Are you taking voltage readings from BIOS or with software in Windows? Does it fluctuate? What about 12v rail?
Just noticed this. "I've simply been raising the FSB multiplier." Er, which one is it? FSB or the multiplier? I think you mean FSB, right? (Aren't most Intel CPUs multiplier locked - can't raise but can lower it?) Try lowering multiplier by couple notches from the original x15 and raise FSB gradually. You should be able to lower your voltage as well. Besides, the gains from tighter RAM timing is minimal. |
|
Back to top |
|
 |
JustJohn

Joined: 18 Oct 2007 Location: Your computer screen
|
Posted: Wed Nov 14, 2007 3:53 pm Post subject: |
|
|
Also, I wouldn't worry about the voltage that much if your temps are fine, and they are. You might be shortening the life a bit, but if it's your backup system that's no biggie. |
|
Back to top |
|
 |
eamo

Joined: 08 Mar 2003 Location: Shepherd's Bush, 1964.
|
Posted: Wed Nov 14, 2007 5:23 pm Post subject: |
|
|
Seoulfinn, yes. I mean the FSB. The CPU multiplier is locked.
Is 3.6 the max? I don't know. I'm guessing I could go higher as long as I can accept the temps.
I'm taking the voltage readings from a couple of Windows software utilities. CPU-Z, Everest and PC Wizard. The voltage reading fluctuates by 0.02 from 1.46 to 1.48 according to CPU-Z.
Lowering the mulitplier? I'll look into that.
Cheers. |
|
Back to top |
|
 |
|