Solved

CPU Temperature Monitoring

I have AMD Phenom II X6 1055t CPU that has a maximum temperature limit of 62 C. I am monotoring CPU Temps with Coretemp and HW Monitor. Both show Core Temps about the same, 22-23 C at idle. But HWMonitor also shows 3 more readings, TMPIN0, TMPIN1, and TMPIN2, which are different than the Core reading of 22 C.

So when I try to keep CPU temp below 62 C, which one do I watch? Core temp (which is 22 c at idle) or other 2 TMPIN0 and TMPIn1. It can not be TMPIN2 because it is already too high at idle, 79 C. TMPIN0 is 7 C higher at 29 C, and TMPIN1 is 11 C higher at 39 C.



So I run IntelBurn tset to load the CPU to 100% and temps increase. According to Coretemp Core is at 52 C, but TMPIN1 is at 61 C. According to core, I am safe since 52 is less than 62 C, but according to HWMonitor TMPIN1, temp is 61 C, only 1 C below max, and I should reduce my clock.

So which temp I watch?

44 answers Last reply Best Answer
More about temperature monitoring
  1. jhaveri said:
    I have AMD Phenom II X6 1055t CPU that has a maximum temperature limit of 62 C. I am monotoring CPU Temps with Coretemp and HW Monitor. Both show Core Temps about the same, 22-23 C at idle. But HWMonitor also shows 3 more readings, TMPIN0, TMPIN1, and TMPIN2, which are different than the Core reading of 22 C.

    So when I try to keep CPU temp below 62 C, which one do I watch? Core temp (which is 22 c at idle) or other 2 TMPIN0 and TMPIn1. It can not be TMPIN2 because it is already too high at idle, 79 C. TMPIN0 is 7 C higher at 29 C, and TMPIN1 is 11 C higher at 39 C. Thanks.

    http://home.gci.net/~jhaveri/Temps.jpg


    TMPIN0 would be the one you want to keep your eye on!! That is your CPU temp. Do NOT use core temps core temps as your guide. I believe TMPIN1 is your mobo.
  2. miwanuma said:
    TMPIN0 would be the one you want to keep your eye on!! That is your CPU temp. Do NOT use core temps core temps as your guide. I believe TMPIN1 is your mobo.


    TEMPIN0 is only 30 C during load, and 29 C at idle (please see the figures above) that can not be correct.
  3. Wich cooler do you have?
  4. Coretemp is the standard for AMD temps. Nuff said.
  5. Conumdrum said:
    Coretemp is the standard for AMD temps. Nuff said.



    Heres a picture of mine.. my TJ max is at 90c.. strange?

  6. ranij93 said:
    Wich cooler do you have?


    Thermaltake SpinQ
  7. Btw the picture i added does not work, and i can't "eddit" the message.
  8. Conumdrum said:
    Coretemp is the standard for AMD temps. Nuff said.


    So when coretemp gives 52 C at load but HWmonitor is 61 C, coretemp is correct, and I am way below the max 62 C for 1055t?
  9. tempin1 is the motherboard chipset temp.im not sure which one though.
    my msi x58pro e hits 81 on tempin1 : )
  10. andrewcutter said:
    tempin1 is the motherboard chipset temp.im not sure which one though.
    my msi x58pro e hits 81 on tempin1 : )


    I need confirmation that CPU temp I should be monitoring is Core in HWMONITOR, not TMPIN0 or TMPIN1. Because if that is correct, I can overclock higher where the Core is at 60 C. On the other hand if the TMPIN1 should be the one I should be monitoring, any higher clock and CPU will exceed the 62 C and possibly toast.
  11. have you tried some other program too like everest ?
  12. andrewcutter said:
    have you tried some other program too like everest ?


    Yes, Speedfan, and it gives the same 3 TMPIN0,1,2 as HWMONITOR.
  13. I you suggest you play it safe, as you know there is always a risk involved. Ultimately it is up to you sir.
  14. jhaveri... coretemp is not giving you tmpin readings at all... its giving you core temp reading... thats all...

    As far as max temps... the Tcase max is 61C and Tctl max is 70... tmpin0 is mobo/system/chasis temps and tmpin1 is your cpu temp...

    AMD always say to trust the bios temps as they can verify that reading only... so check whether the cpu temp in your bios reading is nearest to which temp reading in HWM...

    try Hardware Mania Blackbox... it offers comprehensive monitoring of temps as well as volts... maybe that would help...
  15. abully said:
    jhaveri... coretemp is not giving you tmpin readings at all... its giving you core temp reading... thats all...

    As far as max temps... the Tcase max is 61C and Tctl max is 70... tmpin0 is mobo/system/chasis temps and tmpin1 is your cpu temp...



    Please don't quote Tcase numbers without an explanation, they never do anything except confuse people.

    For AMD processors 62C is nowhere near the max safe temp, so you're fine.

    Tcase is a very specific scenario used in industry testing. It's defined as maximum operation temperature above ambient for a CPU as a set ambient. For Intel standard is 39c, not sure if AMD is the same.

    The actual maximum operating temp(TTV) would be T ambient + T case. In the case of Intel processors, this is 108C. Though the processor throttles at about 105C.

    Tcase is not very meaningful as this maximum is defined under specific conditions temp. Thermal envelope of hardware will change depending upon current temp as thermal resistance varies with temp.

    As for the reason you have diff temp readings. Core temps measure temps inside the CPU, HW monitor and Speedfan take outside surface of CPU temps, hence a difference, though it's usually Core temps giving a higher number, not the other way around.

    In general though, software temp readings always have a degree of error.
  16. banthracis said:


    For AMD processors 62C is nowhere near the max safe temp, so you're fine.



    Thanks banthracis. As abully requested, I checked BIOS reading, and it is same as TMPIN1 in HWMONITOR.

    Which means that at load, my CPU temp is 61 C read by TEMPIN1, and not 52 C - the Core Temp.

    You mention that for AMD, 62C is nowhere near max safe temp. So is the max safe temp higher or lower than 62C.

    If it is higher than 62 C in TMPIN1, I will be OK with this overclock. If lower, I will have to reduce FSB.
  17. All processors cores CPU's/GPU's have TTv >100C.

    Real world, I wouldn't be worried about impact on system life until temps get above mid 80's C.

    Bear in mind while enthusiast's go crazy over low temps, in the PC industry, especially with laptops, temps in the 90's C are considered perfectly normal and fine (very common in Apple laptops).

    Silicon degradation is either a logirithmic or exponential curve (forgot which off top of my head) in regards to temps, so the difference on system life btwn 90c and 100C is greater than the difference btwn say 40C and 90C.
  18. banthracis said:
    All processors cores CPU's/GPU's have TTv >100C.

    Real world, I wouldn't be worried about impact on system life until temps get above mid 80's C.

    Bear in mind while enthusiast's go crazy over low temps, in the PC industry, especially with laptops, temps in the 90's C are considered perfectly normal and fine (very common in Apple laptops).

    Silicon degradation is either a logirithmic or exponential curve (forgot which off top of my head) in regards to temps, so the difference on system life btwn 90c and 100C is greater than the difference btwn say 40C and 90C.


    Thanks again sir. I read so many posts last night on this topic that my head hurts. I was worried about the mention of staying below 55 C (not even 62 C) for AMD Phenom II X6 1055t processor. Based on your advice, I should not worry with my TMPIN1 i.e. CPU temp near 61 C. Actually, I may be able to even increase FSB from current 250 to 285 and try 4 GHz OC. Before I do that, I appreciate if you confirm that your comment is indeed about the AMD 1055t. You being Veteran on the forum, I take your advice very seriously.
  19. banthracis said:
    All processors cores CPU's/GPU's have TTv >100C.

    Real world, I wouldn't be worried about impact on system life until temps get above mid 80's C.

    Bear in mind while enthusiast's go crazy over low temps, in the PC industry, especially with laptops, temps in the 90's C are considered perfectly normal and fine (very common in Apple laptops).

    Silicon degradation is either a logirithmic or exponential curve (forgot which off top of my head) in regards to temps, so the difference on system life btwn 90c and 100C is greater than the difference btwn say 40C and 90C.


    are you trying to reinvent the wheel here..? why would you want to advise some one with your own empty explanation ?

    @OP your processor max safe temp. is 71C
    http://products.amd.com/en-us/DesktopCPUDetail.aspx?id=652&f1=&f2=&f3=&f4=&f5=&f6=&f7=&f8=&f9=&f10=&f11=&f12=

    use HWmonitor for your temps
  20. ortoklaz said:
    are you trying to reinvent the wheel here..? why would you want to advise some one with your own empty explanation ?

    @OP your processor max safe temp. is 71C
    http://products.amd.com/en-us/DesktopCPUDetail.aspx?id=652&f1=&f2=&f3=&f4=&f5=&f6=&f7=&f8=&f9=&f10=&f11=&f12=

    use HWmonitor for your temps


    My CPU is 125 Watts, so max is lower at 62 C.
    http://products.amd.com/en-us/DesktopCPUDetail.aspx?detailId=641&id=641&id=652
  21. as long as you hep it under that number you are more than safe
    Edit; run prime 95 for two hours and report back with screen shot of HWmonitor
  22. ortoklaz said:


    Thanks. So with 62 C max, am I near danger zone at 61 C, and reduce the clock ?
  23. ortoklaz said:
    are you trying to reinvent the wheel here..? why would you want to advise some one with your own empty explanation ?

    @OP your processor max safe temp. is 71C
    http://products.amd.com/en-us/DesktopCPUDetail.aspx?id=652&f1=&f2=&f3=&f4=&f5=&f6=&f7=&f8=&f9=&f10=&f11=&f12=

    use HWmonitor for your temps


    I already defined Tvv in my earlier post.

    Silicon degradation is the term for why CPU's are damaged due to high heat and voltage.

    You once again linked a Tcase number, which for the LAST TIME is wrong.

    AMD is nice enough to define everything. See pg 18 of data sheet
    Tcase 70c, ambinet 48C so Tvv would be 118 C. They even define the changed thermal profile due to their stock HSF on pg 19, which is where the 71c comes from.

    http://support.amd.com/us/Processor_TechDocs/43375.pdf

    Stop spreading false information. Read the data sheet or leave the electrical engineering to the electric engineers.

    edit: Tcase for 125 W CPU's is defined at 44c ambinet, so Tvv would be 105C. See thermal profile 2 in data sheets.
  24. jhaveri said:
    Thanks. So with 62 C max, am I near danger zone at 61 C, and reduce the clock ?


    it won't be "denger" zone but i would try to have my CPU running belove MAX temp. listed by manf.
  25. jhaveri said:
    Thanks. So with 62 C max, am I near danger zone at 61 C, and reduce the clock ?



    Ok, I'll try to clarify this again.

    the number's people have been linking are Tcase numbers, which are NOT maximum operating temps as your software or a thermometer would read.

    thermal profiles vary depending upon ambient and current CPU temperature. As a result, AMD's uses standardized testing to determine a safe maximum operating temperature.

    To determine these numbers, AMD says IF:

    1. Ambient temp is 48C. This is what T ambient is defined as.
    2. Thermal resistance is 0.232 C/W

    then:
    For a 95W CPU, the maximum safe operating temp above ambient is 70C for a 95 W CPU. This is also known a Tcase or the number that everyone likes to link.

    However, this is NOT the temp your software measures. You software, like a room thermometer, measures the total temperature in a given area, it doesn't break it down into Tambient and Tcase.

    If order to figure out the safest operating temperature as you would get from a thermometer, you need Tvv. Tvv is calculated by adding Tcase and the ambient temp so for a 95w CPU, it would be 70C + 48C= 118C.

    For a 95W Phenom processor, 118C is the highest safe operating temperature, assuming 1 and 2 stated above.

    Now, the reason why companies don't give a standard highest operating temp, is because as I mentioned earlier, this number varies depending on temp due because thermal resistance changes as temp goes up and down.

    Hence, companies give a a number based on their testing conditions.

    For a 125w CPU, highest safe operating temperature as your software reads, would be Tcase 61C + 44C ambient=105C.


    So in summary, your current temps are fine, and are over 40C under the max operating temperature.
  26. banthracis said:
    Ok, I'll try to clarify this again.



    Thanks banthracis. I am reading the pdf file you sent, and trying to digest the information. Looks like I am safe and should try to raise the clock, since my goal is 4 GHz. This is my play computer, not the serious data crunching one. So if the CPU fries, it won't bother me, I would learn something. I am running Prime95 for 2 hours as suggested, and report HWMONITOR results later today. Regards, -jhaveri
  27. @ OP don't worry about Tcase temps just stick with core temps,if you trying to OC your chip and getting close to 62C(based on 24/7 operation) core temp#1,2,3,4,5,6..you have to improve your cooling that's all
  28. I think you're just trolling on purpose at this point.

    Hell, even the editors at Tom's here agree that 64C above ambient is perfectly fine for a overclock.
    Quote:
    Cooler Master’s Hyper 212 Plus provided enough cooling for our 4 GHz CPU, the Freezer 7 Pro Rev.2 did not


    bottom of page
    http://www.tomshardware.com/reviews/lga-1156-heatsink,2535-15.html

    Hyper 212 Plus is 64C above ambient per chart here

    http://www.tomshardware.com/reviews/lga-1156-heatsink,2535-14.html

    Granted they were talking about an i7 not a phenom 6 here, but the general gist is correct, as max operating temps for Intel CPU's are about the same as AMD, 108C.

    Some more discussion by the editors here, about how high 60's C is fine, and high 70's C is impressive with an OC. They use an i7 again.

    Quote:
    Let’s look at the processor peak load temperatures. Intel’s cooler results in a constant 68°C when the 2.66 GHz Core i7-920 runs at full load on all four cores. The same scenario results in a 62°C CPU temperature using the Zalman CNPS 10X.

    But it’s the overclocked temperatures that demonstrate most impressively how well the Zalman cooler performs (or how poorly the retail cooler copes with extra heat). The temperature level ranges between 75-79°C for the CNPS 10X, depending on the fan speed, while the Intel cooler is incapable of addressing the additional heat, and the CPU runs at 94°C.

    http://www.tomshardware.com/reviews/amd-phenom-ii-x6-1090t-890fx,2613-13.html
  29. banthracis said:
    I think you're just trolling on purpose at this point.

    Hell, even the editors at Tom's here agree that 64C above ambient is perfectly fine for a overclock.
    Quote:
    Cooler Master’s Hyper 212 Plus provided enough cooling for our 4 GHz CPU, the Freezer 7 Pro Rev.2 did not


    bottom of page
    http://www.tomshardware.com/reviews/lga-1156-heatsink,2535-15.html

    Hyper 212 Plus is 64C above ambient per chart here

    http://www.tomshardware.com/reviews/lga-1156-heatsink,2535-14.html

    Granted they were talking about an i7 not a phenom 6 here, but the general gist is correct, as max operating temps for Intel CPU's are about the same as AMD, 108C.

    Some more discussion by the editors here, about how high 60's C is fine, and high 70's C is impressive with an OC. They use an i7 again.

    Quote:
    Let’s look at the processor peak load temperatures. Intel’s cooler results in a constant 68°C when the 2.66 GHz Core i7-920 runs at full load on all four cores. The same scenario results in a 62°C CPU temperature using the Zalman CNPS 10X.

    But it’s the overclocked temperatures that demonstrate most impressively how well the Zalman cooler performs (or how poorly the retail cooler copes with extra heat). The temperature level ranges between 75-79°C for the CNPS 10X, depending on the fan speed, while the Intel cooler is incapable of addressing the additional heat, and the CPU runs at 94°C.

    http://www.tomshardware.com/reviews/amd-phenom-ii-x6-1090t-890fx,2613-13.html


    dude..why don't you give it up already and bring'n up all this i7 crap,this guy is asking about AMD chip and i believe he just got confused with the temps shown in HWmonitor and want to play safe,besides who wants to run the chip close to " silicon degradation"
    I understand you want to prove your point but why don't you try to do something that's best for him in the long run ,if AMD states max safe core temp is 62C then that's what it is
  30. ortoklaz said:
    @ OP don't worry about Tcase temps just stick with core temps,if you trying to OC your chip and getting close to 62C(based on 24/7 operation) core temp#1,2,3,4,5,6..you have to improve your cooling that's all


    ortoklaz: The prime95 has been running for 2 hours stable, and going on, no errors. Unfortunately HWMonitor froze. I closed and restarted, but it is not responding. Even speedfan is not responding. But before I closed the frozen HWMonitor, the TMPIN1 was 58 C and core temp was 49 C. But the rest of the system is stable. As I am replying this post while Prime95 is still running. This brings me back to my original question. My coretemp is about 9 C lower than TMPIN1 (CPU Temp) in HWMonitor. So do I monitor coretemp (now at 49 c) below 62C or TMPIN1 (was 58 C before freezing)?

    I do understand the confusion between you and banthracis regarding the safe max temp for AMD Phenom II X6 1055t 125 watts CPU. I do appreciate both of you wanting to help - and I am grateful. I am very new to OC, my first one, and relying on expert help on this forum. Thanks again. -jhaveri
  31. ortoklaz: So speedfan is working on my 2nd try - HWMonitor does not. Here is the speedfan readings.
  32. ok if you run prime95 for 2 hours you should be stable,try to restart your pc if you still have problem with HWmonitor uninstall speedfan,hwmonitor and install HWmonitor only again ,if your core temps below 62C you are fine
    http://img218.imageshack.us/img218/8421/55704686.jpg

    Uploaded with ImageShack.us
  33. ortoklaz said:
    ok if you run prime95 for 2 hours you should be stable,try to restart your pc if you still have problem with HWmonitor uninstall speedfan,hwmonitor and install HWmonitor only again ,if your core temps below 62C you are fine
    http://img218.imageshack.us/img218/8421/55704686.jpg

    Uploaded with ImageShack.us


    I did restart and having no problems with HWMonitor. When I see your attached figure, your TMPIN1 is lower than coretemp. As you saw earlier posts, my core is 9 C below TMPIN1, both at idle and load. SO based on waht you say, my core at load was 49 C (not 58 C as in HWMonitor). Does that mean I have a lot more possibililty for OC, since 49 C at load is much less than 62. Please see my idle HWMonitor below and note how TMPIN2 (34 C) is higher than core (23 C).


    I apologize for asking again, but if I go by core which is lower than TEMPIN1 by about 10 C, I have a lot of room to overclock. Is it possible that there is something wrong with my motherboard sensor that the core temp is lower than TMPIN1 i.e. CPU Temp. (Your results looks OK since your CPU Temp is lower than Core temp)
  34. Best answer
    please..drop TMPIN subject ,core temp is the one to watch..if your temp. is 49C under load "core temp" you still have some head room for oc

    when you look at core-temp (application) it won't even show TMPIN

    EDIT;

    SYS = TMPIN0 = Motherboard
    CPU = TMPIN1 = CPU ( CPU heatsing temp, NOT the CPU itself)
    MCH = TMPIN2 = Northbridge

    I'm not 100% sure tho. I have seen TMPIN2 at 128C it could be sensor issue :D
  35. ortoklaz said:
    please..drop TMPIN subject ,core temp is the one to watch..if your temp. is 49C under load "core temp" you still have some head room for oc

    when you look at core-temp (application) it won't even show TMPIN

    EDIT;

    SYS = TMPIN0 = Motherboard
    CPU = TMPIN1 = CPU ( CPU heatsing temp, NOT the CPU itself)
    MCH = TMPIN2 = Northbridge

    I'm not 100% sure tho. I have seen TMPIN2 at 128C it could be sensor issue :D


    TMPIN2 thats what mine says so definitely is..
  36. ortoklaz said:
    please..drop TMPIN subject ,core temp is the one to watch..if your temp. is 49C under load "core temp" you still have some head room for oc

    when you look at core-temp (application) it won't even show TMPIN

    EDIT;

    SYS = TMPIN0 = Motherboard
    CPU = TMPIN1 = CPU ( CPU heatsing temp, NOT the CPU itself)
    MCH = TMPIN2 = Northbridge

    I'm not 100% sure tho. I have seen TMPIN2 at 128C it could be sensor issue :D


    Fantastic, I am so pleased to know that I have a lot of room to overclock. I was worried beacause of other posts, here is the one which made my head hurt with no solid conclusion at the end. It is amusing how much ignorance is out there about coretemp vs TMPINx. Specially look for comments by gareth170 halfway down the first page, here is one of his comment:
    "it is true. core temp give fake readings on phenom II.. you need to use hwmonitor and look at tmpin0 or tmpin1, depending on motherboard.'

    http://forums.overclockers.co.uk/showthread.php?t=18206936

    Thanks again for your help. Tonight I am trying to see if I can get to 4 GHZ and keep coretemp below 60 C.

    Happy Holidays.
  37. Ok just received confirmation from AMD... the core temp max is 61C so it should always be below 62C...

    and TMPIN1 max is 71 so stay below 72...

    END OF STORY...
  38. Best answer selected by jhaveri.
  39. abully said:
    Ok just received confirmation from AMD... the core temp max is 61C so it should always be below 62C...

    and TMPIN1 max is 71 so stay below 72...

    END OF STORY...


    True, End of Story for me. My TMPIN1 is 10 C higher than core. So evrything is consistent. Staying below 61 C for coretemp is same as stying below 71 for TMPIN1.

    Thanks abully for confirming with AMD. Happy Holidays.
  40. The story will never end unless AMD releases an official declaration about this issue, not just a customer service response. Till then I have to trust in "common sense" and a core temp lower than the ambient temp is a nonsense.
  41. @OP..this may help you oc your chip,some good info here;
    http://www.overclock.net/amd-cpus/720502-1055t-owners-club.html
  42. ortoklaz said:
    @OP..this may help you oc your chip,some good info here;
    http://www.overclock.net/amd-cpus/720502-1055t-owners-club.html


    Good info indeed. Thanks.
  43. miwanuma said:

    TMPIN0 would be the one you want to keep your eye on!! That is your CPU temp. Do NOT use core temps core temps as your guide. I believe TMPIN1 is your mobo.




    what should i trust then? LOL
    (sorry for reviving an old thread)
Ask a new question

Read More

Heatsinks Core Temperature Monitor CPUs Overclocking