Solved

Core Temperature

So I'm getting all four cores of my core i7 920 overclocked to 3.8 between 31 and 33 C on idle. VCore is at 1.25. The first core is always the hottest at around 33 C, second core 31 C, third core 33 C, and fourth core at 32. They all fluctuate within themselves by about 1 degree Celsius. My question is whey the fluctuate even that much between each other? Why is the second core cooler than the last core? I thought I put my thermal compound on perfectly - Arctic Silver 5 and I have a Noctua NH-U12P SE2 for a cpu cooler. Is temp fluctuation caused by thermal compound or is there no way to get all cores running at the same temp. I believe the first core will always run a tiny bit hotter, but i'm just wondering. I'm also wondering if I can get my cpu to run under 30 C on idle using my cooler.
9 answers Last reply Best Answer
More about core temperature
  1. There will always be a slight difference between the core temps. Idle temp means nothing at all, run Prime95 or IBT or OCCT and see what your 100% load temps are. They are all you need to be concerned about.
  2. RJR said:
    There will always be a slight difference between the core temps. Idle temp means nothing at all, run Prime95 or IBT or OCCT and see what your 100% load temps are. They are all you need to be concerned about.


    exactly this.
  3. Can't tell much on idle power. At 3.7 Ghz, my son's vcore is 1.125 Use OCCT to torture test and record those values....Prime95 (small FFT's) will be 1 or 2 degrees higher. Also go easy on AS5 till it cures....and yes, 1st core usually hottest and 304C between hottest and coolest.

    http://benchmarkreviews.com/index.php?option=com_content&task=view&id=138&Itemid=1&limit=1&limitstart=5

    Here is the Arctic Silver 5 recommended cure time instruction from the manufacturers web site:

    Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.

    So by my estimation of this statement it would take almost a year of normal use to properly cure the AC5 compound, or almost nine days of continuous power cycles to meet their recommendation
  4. Actually I just reseated my heatsink with Arctic Silver 5 about 3 days ago. XD I read that statement on their site when I wanted to make sure I was doing it right.
  5. I have been using AS5 for years and I have never noticed a significant change in temps after the stuff cured.

    Strings, when you reseated, did you clean the old stuff off before adding new AS5?
  6. Yes, I used the Arctic Silver Thermal material Remover & Surface Purifier
  7. Best answer
    The first question was why the difference in temp between the cores.
    Simply put the cores that are working (running program code) will get hotter than the cores that are NOT working. Generally speaking program code is executed by just one or only a few cores (depending on hyper-threading of the application). Those cores will get hotter while doing so, while the idle cores will stay cooler. Keep in mind that even at idle the OS is running code to do even simple housekeeping functions like checking for mouse movement, keypresses, Interrupt requests etc. etc.. No cpu is truly idle even when idle.
    So IMHO the temp fluctuation has nothing to do with how you applied your cpu compound. Mine does the same thing (I7-920 at stock 2.66gz, no oc)

    As for getting your cpu below a certain temp, there are a couple things you can look at. They are:
    1. Overall air flow through the case. If you're not getting the hot air out of there it raises the over all ambient temp of the components in the case. Think about it this way. If it's 100 degrees outside and your inside house temp is 105 degrees you are not gonna get your house below 100 degrees no matter what (without air conditioning IE liquid cooling).

    2. The cooler itself. Not all cpu coolers are created equal as cooler benchmarks will show you. You'll have to scout around and do you homework to decide for yourself. Personnally I finally settled on a Prolimatech Megahalems CPU Cooler Rev B. but a lot of ppl might disagree and choose something else. What I can say about this is I saw a lot of coolers with some really rough surfaces where they touch down on the cpu and all I can say about that is a flat glassy-as-possible surface will work better than a rough one. The thermal compound should fill microscopic gaps not obvious distortions in the contact point. If you can see grind marks on the face (like one I saw), I would stay away from it.

    3. The fan on the cpu cooler. This is the most likely (and cheapest) thing you can change out that will drop your cpu temps.
    I discovered this recently for myself. With my P.M. cooler I bought a Kaze Jyuni SY1225SL 12SH fan ($10). Long story but I also had a Cooler master 120mm Blue led fan (R4-L2R-20AC-GP at amazon). Last week I decided blue would be cool in the case so I swapped out the Kaze Jyuni for the Cooler master. My cpu temps went UP 10 degrees F.! I immediately took it back out.
    Blue wasn't that cool...

    [edit] PS mine is running at ~24c but again thats without any oc
  8. This is nothing to worry about. Now if you were on a C2Q, and the temperatures on core 1and 2 were different than 3 and 4, your cooler could be not exactly even, but you are running a native quad so no worries.
  9. Best answer selected by Strings.
Ask a new question

Read More

CPUs Core Cooling Thermal Compound Temperature Overclocking