Q9450?

[jcorqian]

I just bought a Q9450, just curious to see what some people out there have OC'd it to? And what their system specs are?

 

[halcyon]

3.4Ghz so far, hoping to take it a little further with better RAM.

 

[mesh1]

jcorqian - what are your stock temps? I just bought mine today also and i put it in my ds3l rev2.0 mobo and my temps seem pretty high with my Noctua NH-U12P.

 

[halcyon]

Are you using RealTemp? If not, try it: http://www.techpowerup.com/downloads/1057/Real_Temp_2.5.html

 

[jcorqian]

I won't have my parts for my build yet until next Tuesday, so I couldn't tell you sorry.

 

[mesh1]

yea im using real temp

 

[mesh1]

I just put my tuniq tower on it to see the difference between the two

 

[halcyon]

Mesh, I've been through the frustration myself. The warmest core on my 9450 chooses to idle @ 46-50C and the coolest 39-44C ...under load the warmest core will hit 66-67C in a ~24C room. I've tried 3 different coolers and 5 applications of AS5. I'm not a rocket scientist but I'm not quite retarded either.

I've read that the DTS sensors on the 9450's are funky and show inaccurate idle temps.

 

[mesh1]

Mesh, I've been through the frustration myself. The warmest core on my 9450 chooses to idle @ 46-50C and the coolest 39-44C ...under load the warmest core will hit 66-67C in a ~24C room. I've tried 3 different coolers and 5 applications of AS5. I'm not a rocket scientist but I'm not quite retarded either.

I've read that the DTS sensors on the 9450's are funky and show inaccurate idle temps.

How funny. Just as im reading through several posts regarding the inaccuracy of these 9450 sensors I come back and read your post. Well, this is most disappointing because it makes it rather difficult to OC / bench. Hmm.. Im curious what Intel has to say about it. I just bought my processor at Fry's this Friday so im still able to exchange it. Did you hold on to your 9450?

 

[halcyon]

Yeah, I did. Using RealTemp it says the DTS sensor shows 28-30C delta to TjMax (the maximum temperature for the processor). I do believe that's close to accurate giving the research I've done. 45nm chip's sensors, as you may have read, become much more inaccurate as the temperature is loaded since that's what the temperature sensors are designed for.

The sensors are not designed to show idle temps, they're designed to tell the processor at what point to throttle back due to excess temperatures. Via reverse calculations are beloved software tries to determine what the idle temperature is (of course ambient temperature calibrations are required for true accuracy). This can all only be accurate if TjMax value is known (and other factors), but since Intel doesn't release that data for its desktop processors its guess-work. It is widely thought that the TjMax value for the 45nm Wolfdales and Yorkfields is 95C...but noone knows for sure.

 

[MooseMuffin]

Those temps match what I'm seeing. I have mine running at 3.2Ghz and the warmest core hits 65C under load, using the tuniq tower with the fan turned most of the way down.

 

[nman729]

3.6 45c load, watercooled on a rampage formula 450x8.

 

[halcyon]

Yeah, I might have do some H20 at some point. I'm truly less than a noob @ that!!!!

 

[nman729]

the h20 220 compact is a great starter kit, and should be more than adiquit for those who dont want to deal with a big bulky air cooler. people told me i would not be satisfied with a kit from swiftech or anywhere, they were totally wrong, i love it, and it was a lot easier than starting from scratch.

 

[halcyon]

the h20 220 compact is a great starter kit, and should be more than adiquit for those who dont want to deal with a big bulky air cooler. people told me i would not be satisfied with a kit from swiftech or anywhere, they were totally wrong, i love it, and it was a lot easier than starting from scratch.

I was just looking at that kit yesterday in MicroCenter. So, you're recommend it to someone that's never done liquid b4?

 

[jcorqian]

I OC'd my Q9450 to 3.8, but brought it back down to 3.45 because I didn't want to push things unnecessarily. Anyway, my temps at this speed seems to be different with each different temp monitor. The hottest core went to about 70 C under Core Temp, 65 C under SpeedFan, and 60 under Real Temp. Do you guys know which one would be the most accurate?

 

[Evilonigiri]

Realtemps reports temps a bit lower than it really is while core temp reports a bit higher. This all varies depending on where the temperature is. Coretemps gets wayy off at the higher temps supposedly.

 

[nman729]

i have a h20 220 apex, thats basicly, the pump, rad res and block seperate. ie for experienced users. i love the kit by the way anyways, with teh 220 compact it has the block and pump all in one unit, and the res and radiator all in one unit. so all you have is the res/rad unit, attached with tubing to the block/pump unit. all you have to do is route the tubing outside the case, and mount the radiator, which is easily done on any case with a 80, 90, or 120mm fan in the back. it will be a little less efficiant than a regular water cooling system, but it will easilly outperform an air cooler.

 

[Gargamel]

You mean you mount it over the exhaust fan at the back of the computer?

 

[Zuwie]

I've got mine at 3ghz without any voltage increase, I use the Abit IP35 Pro board and 2gb ocz reaper 1066. Mine runs very cool even with the stock heatsink. I idle at ~32C in a room temperature at about 22C, and even after hours of torture test, the temps have never gotten above 50C. (this is using real temp)

 

[shadowthor]

hopefully the sticky temp sensors issue is fixed with a new stepping.