[Prong]

I needed 1.375v in bios for vcore which meant 1.34v idle and 1.31v under %100 load in windows after vdroop with LLC disabled. Also needed 1.49v for northbridge and 1.36v for fsb term voltage to get it really stable at 3.6Ghz.

Even I had no problems for nearly a month @3.6Ghz I just reverted back to 3.4Ghz with safer settings. 1.49v seemed just too much for me and I do not have the chance to track nb temp on my Asus P5E3 since it doesn't have a sensor for nb. So I had no idea about nb temp and as far as I knew safe max for nb was 1.45v. I didn't like the idea to run it with a higher voltage than recomended without tracking the nb temp and for now using my q9400 @3.4Ghz for 24/7 use.

Good luck with your oc...

 

[Prong]

Quote:

Originally Posted by redalert I wonder if you had a P45 or x48 board if you would have better luck overclocking the Q9400

Maybe I would with x48. When it comes to P45 a friend of mine with Q9400 and a P45 mobo needs higher voltages even @3.4Ghz. So I don't think P45 would be better luck but maybe x48.

It was/is very stable @3.4Ghz w/1.216v for vcore and 1.29v on nb. This one was another thing that was strange for me. To reach a speed that is only 200mhz more than 3.4Ghz I needed 1.34v for vcore and 1.49v for nb which I thought 1.25/1.26 or something for vcore would do the trick...

If I can have the chance to track my nb temp ( I'l buy a infrared thermometer as soon as possible ) I will again get it to 3.6Ghz and moving up from there maybe I'll try more depending on nb temp...

 

[Prong]

Quote:

Originally Posted by GamingDaemon Well, I did adjust the TJ Max to 100 for the cores (it was 90). The temps now compare properly with what CoreTemp reports. I ran the RealTemp Sensor Test. Here are the results:

Not sure but sensor Core 2 seems to be stuck.

 

[Prong]

Quote:

Originally Posted by redalert I have to push my voltages alot to get past 460fsb(vcore in bios is 1.33375) unless I enable LLC and the only reason I dont keep it enabled is due to temps. I would need to set vcore to 1.4 bios to get to 3.8 stable damn voltage droop and drop is the only thing I dont like with my board. My nb voltage has no effect on my overclock only vcore gets me stable.

Temps are the only reason to keep it enabled or disabled ?

I use enabled LLC only to reach the same speed with lower voltage and my temps doesn't seem to be affected by LLC too much at all. For example my Q9400 is very stable with 1.21v underload @3.4Ghz. With LLC enabled I can set my vcore to 1.24v in bios which gives me only 0.03v for vdroop because of enabled LLC. But if I disable LLC vdroop increases to nearly 0.07v and so I need to set vcore to 1.275v in bios to reach the very same speed.

With both choices it needs to have 1.21v under %100 load. But with enabled LLC I have 1.21v both when idle and underload. With disabled LLC I got to have 1.24v idle which is higher than enabled LLC situation and 1.21v underload.

So if lower voltage means less heat why would you not enable LLC unless you are very close to max vcore ?

 

[Prong]

Quote:

Originally Posted by GamingDaemon Is there anything I can do to fix this? It shouldn't have anything to do with how evenly the themral compound was spread, right?

Are you really willing to see your idle temps ? I would nevermind at all. Important thing is your load temps and it seems to show them to you as temps increases.

 

[Prong]

Quote:

Originally Posted by redalert I have to keep LLC disabled. If I leave LLC on I can drop the vcore down to 1.33125 in bios and be stable at 3.8 but temps hit the high 70's running Linx and running prime95 blend temps are 67-71 thats with an ambient of 21c. The temps are just to high for me and once summer rolls around the temps would be alot worse.

I am sure this is the best choice for you. I believe your mobo does not have much vdroop as mine...Every mobo is different and acts different anyways. To make it clear situation about LLC for me is like below...

Enabled LLC :

vcore in bios 1.24v
vcore in windows idle :1.21v
vcore in windows load:1.21v

Disabled LLC :

vcore in bios 1.275v
vcore in windows idle :1.24v
vcore in windows load:1.21v

So as you might see with disabled LLC I have more voltage on vcore when idle than enabled LLC. Which most probably means more heat. So why not enable it and have the same low vcore both idle and underload which most probably means less heat.

That was what I was trying to say but of course this is what situation is like with my P5E3.

Your mobo might act totally different and it does seem to do so...