Checking to see if OC is typical

[Comdrpopnfresh]

4ghz (9x multi w/ an 1800mhz FSB)
CPU(V): 1.38
HTT(V): 1.32
NB(V): 1.34
SB(V): 1.56

Lower voltages make orthos fail on the large FFTs, though it still seems fine w/ the smaller FFTs. Temps are 39/35 for core1/core2, 45 for the processor, and 31 for the mobo- all idle.

 

[Tempered81]

you should be concerned with Load temps. 4050mhz is a great oc! I had this exact same mobo and cpu at 1.39v 4300mhz.

ram was at 2.1v 880mhz. sb & htt vtt stock and nb at 1.44v. I also had a fan on the nb heatsink.

http://www.xtremesystems.org/f...090&stc=1&d=1233706459

 

[Big Lar]

Your HTT ie VTT is a bit high, otherwise all looks about right.

Larry

 

[BonzaiDuck]

Originally posted by: Big Lar
Your HTT ie VTT is a bit high, otherwise all looks about right.

Larry

The HTT voltage is different from the VTT (CPU_FSB) termination voltage. I was just going to ask him what setting he had for VTT.

On these nVidia chipsets, the recommended voltage setting for VTT is about 1.30V for FSB at or exceeding 400 Mhz, and you say yours is at 450 (1800/4).

Temper this advise with a caution from the Anandtech lab: for Penryn-based cores such as yours, never exceed 1.4V VTT, and try to stay under that limit. For my E8600 with the 780i chipset, it's set to 1.30V and shows 1.28V as reported in Everest or other voltage-monitoring software (as well as BIOS monitor).

What sort of RAM are you using, and what voltage setting for the RAM?

 

[Tempered81]

there is no vtt (fsb term) on that board.

sb & htt vtt stock

there are these options for voltage:
http://www.xbitlabs.com/images...asus-p5n-d/voltage.jpg
the max stable fsb i had was like ~479mhz on an e0 e8400. kinda wished i had a 9.5 or 10 multiplier at that point.

 

[BonzaiDuck]

Originally posted by: jaredpace
there is no vtt (fsb term) on that board.

sb & htt vtt stock

there are these options for voltage:
http://www.xbitlabs.com/images...asus-p5n-d/voltage.jpg
the max stable fsb i had was like ~479mhz on an e0 e8400. kinda wished i had a 9.5 or 10 multiplier at that point.

Sorry, my mistake. I think it's a 750i board, and I would've thought the BIOS would include that feature.

I think I can push my E8600 higher than the current 4.250 Ghz, but the CPU_Z voltage-reading is already at 1.34V idle.

We had some low-end Gigabyte boards that were over-clockable, but you couldn't change the RAM voltage. I wonder if there isn't some custom BIOS version that could be available for the 750i -- probably not. It's a shame, really. But if he can push it as high as he says, I guess it doesn't matter.

 

[Comdrpopnfresh]

I'm using 8 gigs of crucial ballistix tracer @1000mhz, 5-5-5-15-30-2T, 2.01V
I checked my voltages again, here is what they are set to in bios:
VID 1.3875
RAM 2.01
HT(VTT) 1.26
NB 1.34
SB 1.56

 

[Comdrpopnfresh]

Load temps, as reported by speedfan:
Core 0: 43-45C
Core 1: 43-45C
Temp 1(overall cpu?): 56C
Temps 2 (NB?): 30C

 

[BonzaiDuck]

Originally posted by: Comdrpopnfresh
I'm using 8 gigs of crucial ballistix tracer @1060mhz, 5-5-5-15-30-2T, 2.01V

So -- I assume this is 2 x (2x2GB) Tracers?

Be cautious with those. Here's my opinion.

We jumped on these Tracers and Ballistix during 2007 -- lifetime warranty, very over-clockable. At that time, the memory spec was 2.2V. I -- and others -- would push these to settings around 2.175V -- some pushed them above the spec. But even "at-spec," they'd often "blink out" after several months or a year's time. I picked up two sets of rebadged Tracers (NewEgg "LanFest") in late '07, and put one kit in my system. They ran for exactly one year at 2.125V (2.2V spec) and 4,4,4,12 at around DDR=835 Mhz. Those blinked out, and I stuck in the second kit.

I'm guessing that Crucial/Micron was too optimistic with their voltage spec. Maybe the new 2x2GB kits use different parts, but obviously, in your case specifically, the voltage spec has been dropped to 2.0V.

Be careful. If you push the voltage a tad, just remember what I said. Maybe they've been manufacturing these with the same parts and design, and they tightened the voltage spec to reduce RMA's. Maybe you could push them as far as 2.1V if that's the case.

Even so, you're filling all four slots with 'em, so I'd recommend keeping the relatively loose spec latency settings, and running them at spec voltage.

 

[Comdrpopnfresh]

Oh yeah- thanks for the word of advice- I'm running it @ the spec'd epp timings and voltage. I'm more than pleased with the speed @ 1000mhz... though if I try to go for something above 800 with a cas of 4, I'll definitely keep what you said in mind.

was able to turn the SB voltage down to 1.52V

 

[Comdrpopnfresh]

Since someone seemed familiar with this motherboard... how can I change the cpu voltage so it isn't staying at what I have set when the multiplier is dropped? I'm using crystal cpuid to make a three tiered 'speedstep' but it can't change the voltages at all. Even with intel's cpu power saving options turned on in bios, the core voltage doesn't change when the multiplier drops form 9 to 6.

 

[Comdrpopnfresh]

What is VID, exactly? In crystal cpuid, I can change the VID from 1.10-1.250V. When I change it, the vcore readings don't change in speedfan or cpuz, but the VID reading in coretemp does show a change. It seems heaving cpu loading at a high clock speed on my e8400 is more stable with a higher VID