Solved

AMD Phenom II x4 965 Cooling problem

Previously had a pc with no case fans, stock cpu cooler on my CPU and it would record temps up to 80C via speedfan. I had similar results with HWMonitor

I looked into what I could do, and found out a new case, case fans, and new cpu cooler would help me out tremendously.

I have the Coolermaster Hyper 212 EVO CPU Cooler along with 2 case fans, 120mm intake and exhaust. (front and rear)

My cpu right now, (surfing the web and SpeedFan running) is at
Temp1: 39C (shot up to 40C)
Temp2: 56C (wow just shot to 59C)
Temp3: 34C

Running a stress test with Prime95 put me at 62C and counting, I was scared to run it any longer than 5mins.


I re-applied thermal paste after having similar results with the new build, any suggestions?

Previously, the CPU ran so hot on the stock cooler, could that have damaged the chip so much that it is causing this problem?
27 answers Last reply Best Answer
More about amd phenom 965 cooling problem
  1. Grab AOD (AMD Overdrive) and check temps with it.
    Are you overclocking at all ?
  2. Hi swagglikerichie :)

    Your temps are not excessive on AIR cooling and depends on the OC or core voltage applied.
    Prime95 is an aggressive tester although Temps are not bad at 62C and again depending on how long you test for and which test you do.

    There are better alternatives. Download AIDA64 which can also test your sub systems, DIMMs,GPU etc.

    Do not allow your system to exceed 62c under max load and you will be OK.
  3. While gaming cpu hovered around 65C, but the gpu (AMD Radeon HD 7700) was at 50C and one of the fans started to whistle a little bit. I believe it was the CPU fan because I haven't heard anything like that before.
  4. Unless you can get exact temperature, you'll never know and that you can't unless you use proper program, AOD.
  5. Ok I installed Overdrive and it has been reporting the same temperatures that SpeedFan has at idle. Should I test under load and get back you?
  6. Yes, do that. Turn Logging on and use AOD tests under full load. Idle temps are not that important.
    Don't forget that AOD uses temperature margins so higher that number is, cooler it is. Also shows core temperatures which are even more important than package temps most other programs show.
    Goal is to hold temp margins above 10c.
  7. I ran AOD and the thermal margin remained at 39C, under prime95. SpeedFan fand AOD reported the same temperatures whilst looking at "Fan Control" section of AOD.

    They both shot up to almost 69C
  8. MeanMachine41 said:
    Hi swagglikerichie :)

    Your temps are not excessive on AIR cooling and depends on the OC or core voltage applied.
    Prime95 is an aggressive tester although Temps are not bad at 62C and again depending on how long you test for and which test you do.

    There are better alternatives. Download AIDA64 which can also test your sub systems, DIMMs,GPU etc.

    Do not allow your system to exceed 80c under max load and you will be OK.


    The maximum operation temperature for the 965 is 63C, 80C is far too warm.
  9. That's what I've been saying, but user CountMike says it is the thermal margin that is the most important
  10. Thermal margin is the difference from the max temp, so in this case the difference from 63C.
    I personally use coretemp, it has been accurate with my 965 (OCed to 4.2), reporting same temps as AOD.
  11. Gam3r01 said:
    MeanMachine41 said:
    Hi swagglikerichie :)

    Your temps are not excessive on AIR cooling and depends on the OC or core voltage applied.
    Prime95 is an aggressive tester although Temps are not bad at 62C and again depending on how long you test for and which test you do.

    There are better alternatives. Download AIDA64 which can also test your sub systems, DIMMs,GPU etc.

    Do not allow your system to exceed 80c under max load and you will be OK.


    The maximum operation temperature for the 965 is 63C, 80C is far too warm.


    Yea thanks for picking up on that Gam3r01. I forgot TJMax is at 62C for that chip.
  12. CoreTemp isn't accurate, it just said 31C idle or not lol

    I thought a reason that HWMonitor or CoreTemp didn't work was because phenom processor's weren't able to be read, and you have to look at the temperature of the motherboard.
    HWMonitor remains at 31C as well under stress test
  13. There could be a faulty temp sensor in your processor, if that many programs cant read it properly.
    I personally dont trust speedtemp either.

    If the system is not throttling or shutting down technically you would be ok.
    Post a pic (if you can) of AOD while under load
  14. Gam3r01 said:
    There could be a faulty temp sensor in your processor, if that many programs cant read it properly.
    I personally dont trust speedtemp either.

    If the system is not throttling or shutting down technically you would be ok.
    Post a pic (if you can) of AOD while under load


    What should the picture of AOD be? "Fan Control" section or "CPU Status"
    - My gut says go with the obvious lol
  15. CPU status lol, the portion with the graphs and thermal margins
  16. Ok, under load AOD and HWMonitor "Fan Control":
    http://imgur.com/EvbsDmY

    Under load AOD & HWMonitor "CPU Status":
    http://imgur.com/N26MMko

    Idle "Fan Control":
    http://imgur.com/N26MMko
  17. So does your thermal margin always report 39C or does it ever change?
  18. Never changed, idle nor stressed
  19. Best answer
    I think your temperature probe is non functional.
    If your system is stable, does not crash, and does not throttle I would not worry about it.
    A 212 EVO (even with 2 case fans) is more than enough for a 965@3.8GHz
    My EVO managed to handle 1.5v easily.
  20. Throttle eh...

    Every now and then it'll "hiccup" for about 45 seconds until it "collects" itself all over again, then it goes back to normal.

    The rig I had before this contained (embarrassingly) a case with no intake nor exhaust, and stock cpu cooler.

    Could the countless hours and hot summers have damaged the chip, mobo, or both?
  21. I doubt there was any damage done to the system, they usually would shut down to prevent this.

    If you watch the clock speeds when the hiccups happen see if they drop.
  22. swagglikerichie said:
    Ok, under load AOD and HWMonitor "Fan Control":
    http://imgur.com/EvbsDmY

    Under load AOD & HWMonitor "CPU Status":
    http://imgur.com/N26MMko

    Idle "Fan Control":
    http://imgur.com/N26MMko




    Little bit concerned on the 5V rail at just 3.027 Max and your 12V rail is 13.992.
    This maybe due to using two Monitoring systems at the same time.
    If it continues I would check out the PSU.
  23. Oh nice catch there.

    Check the voltages in the BIOS and post them here, might be the reason you are having issues.
  24. ok here they come:
    https://www.dropbox.com/s/m37rr66bma82481/2016-08-16%2022.18.34.jpg?dl=0


    I'm thinking about changing the bios "mode" to standard mode

    Quick update: Thermal Margin idle or load in AOD now reads 70C
  25. swagglikerichie said:

    I'm thinking about changing the bios "mode" to standard mode


    I did so and now the thermal margins are successfully updating in real time
  26. What are they reading now?
    Odd the BIOS mode made a difference.
  27. Gam3r01 said:
    What are they reading now?
    Odd the BIOS mode made a difference.


    Made a new thread for the new problem haha

    http://www.tomshardware.com/answers/id-3157225/amd-phenom-965-mhz-questions.html
Ask a new question

Read More

Cooling Cases CPUs Phenom