Sign in with
Sign up | Sign in
Your question

I7 2600K water cooling temp questions?

Last response: in Overclocking
Share
May 13, 2012 7:54:55 AM

Hi all,

I just finished my new build.

The components i used are;
I7 2600K OC 4.7 GHZ
MSI P67-GD65

XSPC rad RX480 with 4 push fans
XSPC rad RX240 with 2 push fans
XSPC pump/res 750l p/m
XSPC Rasa waterblock

When i run Prime 95 (torture test, Min FTT 8k, Max FTT 8k, FFTs checked on)
Im getting the following temps;
Core1 min 20 max 61
Core2 min 20 max 65
Core3 min 22 max 68
Core4 min 21 max 63

At 4.7 GHZ CPU-Z is showing 1.424V

My questions are;

Are these good temps?
- how about the temp difference between cores?
- I used artic silver 5, and applied it in 1 line in the middle is this the correct way?
- Prime 95 (torture test, Min FTT 8k, Max FTT 8k, FFTs checked on) is this the best test method?

Thanks,

Tammo
May 13, 2012 11:54:37 AM

I havent tried prime95 on my 2600k, but gaming temps are in the low 50s reported by CPU-z.

Clock is at 4.8 and cooler is the corsair H100 kit.
May 13, 2012 11:59:39 AM

Well those temps are a little high because your voltage is so high. For a sandy bridge CPU, i doubt you need 1.424v for 4.7ghz. I used to run 4.5ghz on my 2500k at about 1.3 volts.
Related resources
May 13, 2012 2:02:50 PM

in bios i have set volts to 1.350v but in cpu-z its showing 1.424 volts under 100% load.
Is it possible that i did not apply the thermal paste correctly?
May 13, 2012 2:15:45 PM

No, the temps make sense, and thermal paste has nothing to do with voltage.
Are you using offset voltage or vcore?
May 13, 2012 4:02:28 PM

Ok, are you sure thats the lowest stable voltage?
Also I guess you should reapply paste and make sure your loop is 100 percemt functional, as having the same temps is strange. Do you also have a gpu in the loop?
May 13, 2012 4:32:43 PM

yep untill now the lowest stable voltage.

i just runned intel burn test at high;
results

core 0 min 27 max 58
core 1 min 24 max64
core 2 min 25 max 67
core 3 min 25 max 63

what method do you suggest for applying thermal paste?
cant it be that it needs like 300 hrs to spread? there was something about this in the manual of the thermal paste
May 13, 2012 4:37:28 PM

Well those temps arn't what I would call bad, especially at over 1.4 volts, but overall the voltage and temps arn't great.
1. Why is the bios and OS showing two different core voltages?
2. Is the correct value 1.3 or 1.4?

With thermal paste you want the center area of the cpu (the die) to be covered with the thinnest possible layer of paste. After you put on your block, press it down and twist to make sure it is spread out. (your block is on tight right?)
May 13, 2012 5:03:53 PM

with the last test i had coretemp on.
It showed 1.3711 volt.
why there is a difference between bios and software i don't know.

I used Arctic silver 5
On there website they recommend the line method, i did it aswell.
They also say:
Arctic Silver 5:
Break-in period: 200 hours (Break-in period will occur during normal use.)
Temps will drop several degrees over the break-in period measured with a thermal
diode in the hottest part of the CPU core.
May 13, 2012 5:10:11 PM

Hmm. Makes sense. What air cooler were you using before?
May 13, 2012 5:13:41 PM

scythe mugen 2 rev. b in an antec 900 case with 10 fans and 1x 200mm fan :D 
Im partly deaf now ;)  thats why i wanted water cooling.
May 13, 2012 6:17:16 PM

Oh hey, same cooler I'm using now on my 2500k. I was seeing around 60 degrees at 4.5ghz @1.30 volts.
I guess your temps should be better, especially with 2 rads... I honestly dont know what to tell you....
May 13, 2012 6:35:14 PM

thats ok m8 i appreciate the help.

I just did a prime95 blend test and the highest temperature was 61 degrees.
Which test do you use?

I think ill use it for a week and then ill do some test again maybe the temps will be lower by then.
May 13, 2012 7:47:47 PM

I also use prime. I mean, 61 degrees isnt bad or anything, it just seems with your setup it could be better. Then again, maybe I dont know what Im talking about, I;ve never clocked a cpu that high...
August 16, 2012 8:48:19 AM

:hello:  I just saw your post and I too, some time ago was in the same predicament. Now what I am about to say might not be the root of your issue(s) but it's worth looking into. You have to factor a few things when trying to figure out those temps.
1. If two persons currently have the same room temps but one of those persons live in a humid and hot zone and is having that type of day while the other person lives in a normal temp zone and is having a cool day with the same temps then both their temps are going to be different in their liquid cooling. So even if you see someone else's temps (online) lower than yours, you have to facture their current zone temp activity.
A friend of mine (lives in a hot zone) and I (live in a cold zone). We have done this test a few times during the year (spring, summer & fall) and we have proven this fact.
2. You must remember that your room temps are going to affect your cooling temps. To prove the impact this has try and do this test. If in the night at 10 p.m your room temps are for e.g. 75'F, run your stress test and take your results. Now if you can get up in the morning when your room temps are the same 75'F, then do the same test exactly as you did the night before. You will see that the results aren't the same :bounce:  .
3. The other major factor that you must consider is if:
a. Is there proper air movement in your computer?
b. Are your radiator fans good static pressure fans?
c. Given that radiators are design differently to operate either with low, mid or high rpm fans, are the fans that you are using on your radiator, the most suitable one for it?
Previously I had some regular high cfm low dba fans and I thought all was well with my setup :non: , then I learned about static pressure and its importance. I replaced the original fans with the static pressure fans and saw an immediate 5 degree difference with my setup. I currently have a 2600k and have clocked it as high as 5 GHz (only in the winter lol), 4 GHz in the summer.
Another thing you might want to look at is the pump. In my setup I get better temps when I have a higher flow. I have one of those swiftech pumps (650 I think) and it has 5 settings from low to high.
I hope these thoughts help.
The dude
a c 235 K Overclocking
August 16, 2012 9:14:39 AM

tammo007 said:
Vcore

I used this tutorial a year ago

http://www.youtube.com/watch?v=n7mHYCd1Jho&nomobile=1

Before watercooling i used air cooling.
The temps were about the same as now with watercooling.
i find this a bit strange


I watched that overclocking tutorial, and it is very, what's the word, uh, lame.

There's been a lot learned since that tutorial was posted a year ago, I suggest you totally manually overclock through your bios to take better control of your operating system voltages, which you are using too much of for a 2600K to run 4.7ghz.

Find yourself a manual overclocking guide and use it instead.
http://www.tomshardware.com/forum/265056-29-2600k-2500k-overclocking-guide
This guide was also done using an MSI motherboard (same model as yours), and a 2600K, but shamefully has not been updated even once, since it was originally posted, but it will still give you a better understanding of manually overclocking the 2600K.

By the way, just because you are water cooling now does not mean you can do away with cooling fans, unless you are also water cooling your motherboards voltage regulators (Essential to keep cool), and your memory modules, you still need airflow over them to keep them cool.

Also with a standard water cooling loop you're probably seeing idle temps about 3c above your ambient room temperature and with that much used voltage to reach 4.7ghz your load temps look about right, most expect record breaking results going from air to water cooling, but the Sandy Bridge CPU laughs at that expectation, you're still at the mercy of ambient room temperature.

Taking more control over your voltage will yield lower load temps out in the operating system, however you're not having any crashing issues now are you? so why the worry?

Most of us that have used Artic Silver 5 spread out a very thin, tissue paper thin even layer of AS5 then mount the heatsink, if you're questioning your application method, pull the water block and inspect the thermal footprint.

The worst thing you can do regarding thermal compound of any kind is use too much of it, you are only filling the microscopic imperfections between the 2 mating surfaces and it takes very little to do that.

Thermal compound is a heat conductor, but if you use too much of it, it can become a heat insulator, so pull the water block and inspect what you've done, that way you'll be 100% sure and need not question the application.

You can also use My Guide it is a more aggressive guide disabling literally all Intels features except Thermal Throttling, and Hyper Threading, but many have reported back to me using the guide yielded lower voltage use in the OP/SYS and lower load temps, but it's bios terminology is based off of an ASRock motherboard, and that may be a problem for you.

It's also been updated quite a few times since it was originally posted.
!