My Q6600 Results

Da Man

Gawd
Joined
May 1, 2001
Messages
724
Figure I may as well post since everyone else is :p

My system
Q6600
asus p5b delux
Thermalright Ultra 120 extreme with a SPF fan (68ish CFM one)

Evga GeForce 8800 GTX
4 gigs of the cheaper g.skill 800 mhz DDR2 (the one rated at worse timings)
2 raptor 150 gig hard drives in raid 0
4 samsung 500 gig hard drives in raid 5

Corsair 520 watt PSU


Started out with retail HSF. With ~20C ambient.. core 1 and 2 would be ~40C. core 3 and 4 would be ~50C idle.. and a rise of about 20C for each under load. Fairly hot but was 100% stable.

Threw on the Thermalright. Bottom seemed very machined, wasn't happy with it, but tried it anyway. Same temps as with retail give or take a degree. Knew that this wasn't right.. so I went ahead and lapped the Thermalright ultra 120 extreme down to the copper.. then polished it off. Best sandpaper I could find at lowes was 600 grit, so that is as far as I could go (did the base sanding with 320 I believe).

Huge huge drop in temps.. maybe 35c idle for core 1 and 2, and 40C idle for core 3 and 4. Load brought them up by around 20C.

Dropped ram to slowest setting and have been alternating raising FSB with raising vCore if I started crashing. Right now I am at 350 FSB (3.15 GHz), with ram at I believe 700 MHz. Going to let it burn in a little bit more then see how much higher I can go. IDLE is around 38/43C (depending on which core you look at). Load is around 63/53C. Happy with those temps.. 63C isn't that horrible.

2 concerns at the moment

1) CPU 1 and 2 are always 5C higher than 3 and 4.. I think it is just a "worse" chip.. so I will be limited by that.. as the worst chip in the pair will die first.

2) My PSU may not be able to deal with that much overclocking. I have the vCore right now set to 1.4.. but it boots up at 1.35 in OCCT. Under full load this drops to 1.32. I need to see from you guys.. is this drop in volt typical.. and due to the quality/specs of the motherboard? Or is my PSU starting to max out.

I COULD buy the 620 watt PSU to get a little more speed if that is the bottleneck.. but not really worth the extra $100 for me. My goal was to make it to 3.0 GHz, and I definately can do that.. so I may just back it down to that and use this PSU.
 
Your PSU isn't holding you back.

What are you using to measure temps? You're idling with a quad lower than I'm idling with a duo on water :confused:
 
MM using speedfan 4.32.

Dropped CPU back down to 3.0 GHz, 1.3V, and bumped up my ram to 1.0 GHz. With these settings, my vcore does not budge under load.. which is good.. I was not a fan of my vcore dancing around with my clockspeed.

Looks like my idles after a few minutes of use (i.e. case warms up a little).. are 35C/40C. Under load goes up to 55/60C. Other temps were a cold cold boot.

GPU still at clock.. in vista.. got a 3d mark 06 score of 11,833. Sound about right? Not sure how much I can OC the GPU... not going to mod it or anything.. so maybe not worth the effort.
 
Burned in all night a 3 GHz 1.4V .. ram at 666 MHz 2V

Vcore:
my.php


cpu0 temp:
my.php


cpu1 temp:
my.php


cpu2 temp:
my.php


cpu3 temp:
my.php


5v line:
my.php


12v line:
my.php
 
speedfan

i am using the speedfan and coretemp

but my loaded temp is like 60 degressC on all core..(58-63

but in coretemp it is like 15-20 degress higher..( not reaching 80 degressC yet)

my friends told me to trust the higher number...

i have been asking the same question in different forum.but not much reply...
don't know where is the real number

cpu is running at 1.42v 355fsb 3221mhz Q6600)
with zalman 9500
 
Well I would trust the intel temp utility #1.. which I think speedfan matches. Humm
 
Sounds like either core temp or speefan is bugged.

From the core temp faq

Intel defines a certain Tjunction temperature for the processor. In the case of Yonah it is 85C° or 100C°. First of all the program reads from a Model Specific Register (or MSR), and detects the Tjunction temperature. A different MSR contains the temperature data, this data is represented as Delta in C° between current temperature and Tjunction.

So... they are both getting the same # from the core. One is assuming the max is say.. 100C.. The other is assuming the max is 90C. This is why you get a difference of 10C in the readings.

Would have to know what setting each is using for the Tjunction.. and what the proper Tjunction is..

Either could be right.. tbh speedfan seems more realistic but who knows.
 
First, what did you setup in OCCT to get it to read your temp data? I have tried and tried but nothing ever worked for my P5B-Del and Q6600.

My q6600 had some major temp difference (0/1 were about 6-7C hotter than 2/3). Long story short, I believe this was due to a concave and uneven IHS. After I lapped the CPU, the delta t between cores dropped down to about 3 which is much better.

My advice to you is to lap your CPU and heatsink. I just posted my results:
Lapped my q6600 (pics and temp results)
Lapped my Ultra-120 X (pics and temp results)
 
My heatsink is already lapped.

Did think of lapping CPU side, I thought they were pretty precise.

To get temps in OCCT.. open up speedfan, THEN open up OCCT
 
My heatsink is already lapped.

Did think of lapping CPU side, I thought they were pretty precise.

To get temps in OCCT.. open up speedfan, THEN open up OCCT
Did you take any pics? How did you apply your thermal compound?
 
No pics I was in a hurry.. it was a very flat finish on the heatsink. Sanded through all the zinc so it was all copper, then smoothed it out up to 600 grit sandpaper (need to find higher for next time).

Thermal compound I am pretty good at applying.. did paper thin layer smoothed out with a piece of plastic.
 
Back
Top