Very nice little writeup. Evened out my core temps nicely. Thanks Redcorn!!
Printable View
Very nice little writeup. Evened out my core temps nicely. Thanks Redcorn!!
Ive done this procedure on two other installed quad cores and they had even temps all across. Both had to install the back plate though and no springs we hard mounted them.Quote:
Originally Posted by Sideroxylon
Why run large ffts ? small ffts can give 4-8C higher load temps even with priority set at default
Also different boards seems to recognise the order for cpu0 to cpu3 for quad cores differently from what i tested with Q6600 on intel d975xbx badaxe vs p5b deluxe and asus 680i striker.
interesting nonetheless... will try without springs and see as my first die seems to be always hotter
this with swiftech g4 storm stock mount
http://fileshosts.com/intel/Asus/P5B...ime95_load.png
http://fileshosts.com/intel/Asus/P5B...s_24hrmark.png
Well I don't have the back plate to install but I got all my cores within 2c of each other. I was somewhere between 3-6c difference between cores before.Quote:
Originally Posted by redcorn
Now I'm wondering if I should remove the springs and hard mount it.
2c difference between the cores with the springs installed is not to bad, I guess if you didnt want to mess with it anymore just stop there until you put a back plate on later.Quote:
Originally Posted by Sideroxylon
Man that 975 chipset can really push the quad core well. I wish the 680i chipsets could go past 300 fsb. I wish intel would do a 975 chipset with nvidia SLI. Here is my small TTFs which is probably the equivalent to what TAT does to increase temps fast. This is at 78F or 26C ambient temps with fan on 5/10 and pump half power. One hour on prime 95 small TTF
http://img407.imageshack.us/img407/5223/smallttfsl9.jpg
IM SORRY FOR RESURECTING THIS THREAD!
But i think its a very good thread, and should be bumped back up to the top.
Thanks for this, it's interesting :)
I will try this soon as my temps can vary up to 10C between the hottest and coldest core.
i so needed this hardmount it is!!
And if you go here it shows the Kentsfield quads oriented differently,
http://tweakers.net/reviews/614/18
This is somewhat confusing, not the procedure, just the initial orientation of the waterblock, by the way Redcorn, great guide, should be in the FAQ section.
Utnorris
hardmounted my fuzion and all cores are 1c of eachother :)
Oh yea?
Sounds logical...
Ill try and even out my 6-8ºC Difference...
any pics of the hardmounted fuzion?
Hey, great guide, I can tell it took a LONG time to get it all done, :clap:. I spent the last 2 hours trying to duplicate your results, and have not had success. I've re-mounted my Apogee GT, removed the little plastic spacers, and after a LONG time trying to get it right, I don't think I can with my proc. I took my temps form something erratic and 8C difference to:
C0-54
C1-54
C3-49
C4-49
This is with the processor under load at 3.6ghz with a lapped IHS, and BOWED apogee GT.
The first 2 and bottom two tend to stay close to each other no matter what (this is after 2 re-applications of Arctic Silver), so I'm guessing that one core is just a bit of a dud. Either way, I shaved off 3C off with this whole exercise, and stabilized my temps some, so again good job. If you have any advice for my setup, i'll gladly accept it.
Redcorn, great guide. Just spent about an hour and a half remounting my apogee gtx to my q6600. I used liquid metal pro, so it took me forever to clean everything off. I actually had to use a no-scratch scotch-brite pad and turtle was chrome polish, followed by articclean. Load temps with my q6600 B3 are more uniform across the board 54,54,54,52 with ambient of 72.2 F. No matter what I do, my idle temps have cores 0, and 2 higher. Just must be the processor, cause it's done this with many remounts with any TIM.
Anyone looked it the D-tek pro-mount kit?? I get a 10c split across my cores on my G0 which i can't seem to get down, but i managed to get them within a 1-2c on my B2 using the same technique