Hello all

Just today I installed my new Arctic cooling freezer 7 pro on my Q6600. The installation was simply too easy=P.

I overclocked my Q6600 from 2.4 to 3.0 and I'm at 58, 58, 55, 55(Coretemp reading) at load after 2 hours on Prime95. My question is, I felt like my line of AS5 was a bit thinner on the top compared to the bottom of the CPU(I did the 1 line method for Quad cores). As you can see with my temps, it looks like two cores are running 3-4 degrees above the other 2 cores. I'm wondering if that smaller line of AS5 at the top has caused this. Is it worth taking everything off again to possibly correct the thinner area of line? Or should I just be happy I went from 68 load to 58 load with this 30 dollar purchase?=P

If I redid the AS5 and got the first 2 cores to 55 load as well I could probably move to 3.2 and stay under 60C so I'm debating whether it's worth it.

One other question, with SpeedFan I'm reading 34 ambient, that seems high compared to most that I've read on this forum(22-24). Should I put in a few more fans? Or could I just have a hotter room(set to 74-75).


Thanks in advance-

Variations in temperatures of the cores is very normal and common. I don't think you will ever be able to get all the cores at the same temps. I think the variations in temps are more to do with core loads. Be happy with the temps. If you can go to 3.2Ghz without any voltage bump, the temps should not increase by much at all.

Also, thermal paste takes time to set. After about 300hours, your temps should drop 1-2 degrees lower.

Omg 300 hours are you seriously serious?

not 300 hours.it say under normal use(6hours) takes about 10 days to set.but if you prime it none stop i reckon 4 days the thermal paste will set due to 100% 24/7 compare to normal use of average 40% load for media/internet user or 80% load for game user.

Quads will have 2 cores that show being hot, they use 2 diffrent temp readings, Tcase, and t junction if i remember correctly. Anywho, this is perfectly normal for a quad core, plus you have to remeber, 2 cores are sitting next to each other, and 2 others on the outside not surrounded by cores. Also 1 core is going to be consistantly hotter because it is the one getting the most of the workload with singlethreaded apps. anywho check out this link itll explain it waaay indepth .
http://www.tomshardware.com/forum/ [...] ture-guide

If you keep your computer on 24/7, 300hours is only 12.5 days.

http://www.zalman.co.kr/ENG/produc [...] sp?idx=221

for the zalman's high performance thermal paste it says 7days of normal use.so if 100% load it will take much shorter for it to burn in.

and if the break in period mention by the Artic Silver 5 is only 200 hours.so thats not the 300 he mentioned.and actually break in is when the themal paste is heated up and then cool down to set into the metals on the heatsink and IHS.

now you get it?

So it appears that a 3-4C degree difference between cores is somewhat normal. So pretty much everyone is in agreement that if I reseated, I could possibly not even have a change in temperatures and could have set it properly the first time.

I just found it strange that Core#0 and Core#1 are both the same temperature always at load, and Core#2 and #3 are both the same at load just 3-4C less than Core#0/1. It was my first time applying AS5 and I didn't do a test run with it because the tube was so small and there was definitely less at the top of my line than at the bottom. I was thinking the dies were set together(Core#0/1 together Core#2/3 together therefore resulting in higher temps on Core #0/1 because the smaller amount of AS5).

Perhaps I'm just seriously paranoid because it was just "too" easy to install this heatsink/paste. I've now been stable at 3.0 with no vcore increase for 4 hours now on Prime95. Still at 58-58-55-55 temps.

Thanks for all the help and if anyone else thinks differently and that re-seating could result in 3-4C temperature change on Core#0 and #1 let me know.

Great... failed prime95 on core#0 at 5 hours 7 minutes lol. Guess I'll up the vcore one notch.

you need to up the vcore by 2 notch but you may get away with just 1 notch.the temp seems fine.go and shoot for 3.2ghz.but the vcore jump and temp jump will be inporportionate to the speed increase.

The temperature variation could also be attributed to the convex shape of Intel IHSs. I would leave well-enough alone if I were you. All your temps are safe.

Hi , I'm running the same HSF as you, and with the settings in my Sig, i run @ idle 41/41/44/44 with the fan set @ 100% through Speedfan, and @ load i get 54/54/58/58C, and that's @ stock voltage. the only thing I upped was the NB Voltage.

I run this with an ambient room temp of about 24-28C. I run a HD3870 XF setup so IMHO I got good temps.

BTW Its normal to have different temps, and has ppl suggested, there are reasons for this. The first and mostly proven is the fact that the IHS are in fact either convex or concave, but mostly convex. Lapping may help get more equal temps, but will also void you warranty.

The other reason mentioned is the fact that sometimes you can get "stuck" idle sensors, since the sensors are designed to warn when you get to the upper limit of the Tj, and not to directly monitor the actual temps.

There's a Core2 Temp guide in here somewhere that addresses all the specifics regarding OC'ing temps and whatnot. Linky->http://www.tomshardware.com/forum/ [...] ture-guide

Hope this helps.

On the Arctic Silver topic, I believe that it's 300 total hours to set in, including the time that the computer is off. Check out the product descriptions on NewEgg - the paste needs the whole turning on/off process to "break in" correctly.

From the Core 2 Quad and Duo Temperature Guide - http://www.tomshardware.com/forum/ [...] ture-guide


Section 9: Calibrations

Note 2: It is preferred that Idle and Load tests be conducted as close to 22c Standard Ambient as possible to allow for a normal temperature ceiling, and to maintain environmental consistency for more uniform comparisons among C2Q / C2D variants and system platforms.

(A) Measure Ambient near computer case air intake, clear of warm exhaust. A trusted analog or digital thermometer will suffice.


Section 11: Offsets

Tips:

(D) Graphics Cards which display a sensor labeled Ambient, must not be used for measuring room temperature.


Comp

Another reason for the difference between temps may come from the fact, that a Core 2 Quad is made from two Core 2 Duo dies. That's why core#0 and core#1 have the same temp (one die) and core#2 and #3 are a little hotter (second die).

On my Q6600 (OC'd to 3.6 GHz0 with a TRU120E) with 4 instances of Prime95 running, Cores 0 and 3 hit 65 C and Cores 1 and 2 hit 60 C (??). Hand lapped the base of the HSF to 1500 grit. Did two things - flattened the base and exposed the copper core. Did not lap the CPU IHS.

Took a different approach with the AS5. I divided the top of the IHS in quadrants and put a grain of rice dab in the center of each one.

i did actually mention it if you read carefully.