20

HWMonitor Values

Are these temperature values below the permissible level?

  • 16
    I have to recommend that you try some additional tools to make sure your sensors are being read correctly. most systems will shutdown if the CPU temp exceeds 100C, and sensors notoriously give output that is difficult to read if the software doesn't know them. I usually use CoreTemp, CPUZ, and SpeedFan. – Frank Thomas Aug 14 at 8:50
  • 2
    Those are not really legitimate core temperatures. Your system wouldn't even run at those temperatures. Those are also very close to the temperature if the software was not reading the sensor incorrect (based on the limits of the value of an integer-based variable) – Ramhound Aug 14 at 9:35
  • 6
    115°C is crazy high, yes. That said, it's known that some AMD CPUs apply an offset to the temperature they report vs. their actual temperature, and their mechanisms for temperature reporting in general seem to be less accessible than on Intel CPUs. I think there is a fair change your software is applying an incorrect offest, especially as a genuine 50°C difference between core and package temperatures seems unlikely. That said, if this is under load, I suspect those package temperatures might be low; I would expect to see more like 80°C under load on air. – Matthew Aug 14 at 16:32
  • 1
    Note that many of these read-out tools mess up the F/C conversions, or simply have bad math in them. Try at least two different tools to verify if these numbers are real. – Aganju Aug 14 at 19:12
  • 1
    Online searching indicates that the maximum operating temperature is 102C, and after that it will shut down, so this reading is probably wrong. – Nonny Moose Aug 14 at 22:41
39

The max temperature for many CPUs is listed in the 105-110°C range. But for long-term use, you’re much better off keeping things below 80°C in general and only pushing up to 85°C at the most

To make sure your software is giving you an accurate reading you could compare the results with those of another temperature monitoring application. There are several free lightweight tools to monitor your CPU temperature, one of which is Core Temp.

It would be advisable to check your heat sink to see if it has been dislodged, or that the thermal paste is still connecting the heat sink to the CPU, as running your CPU at that temp is sure to cause damage in the long run.

The GPU temperatures that are shown in your question are not in the danger zone.

| improve this answer | |
  • As hinted, the model of the CPU matters. My ~10 year old CPU in my desktop shouldn't go above 100 but my mac's cpu said something like 110 under stress (compiling a big java project) for a while so I was nervous, but yeah, apparently that was expected, whatever number it was (it may not have been 110) was within the acceptable range for peak (probably not extended use). – Captain Man Aug 14 at 19:00
  • At the startup the temperatures range between 90-95°C. So it is generally high!!. CPU is AMD A10 – user41855 Aug 15 at 4:39
  • From my experience, I think most modern CPU can work for many years even when running more than 90C most of the time; I used to own a HP DV6-6001TX with Sandy Bridge i7, and because the fan went kaput (I just used a blower fan as a fix), most of the time it runs more than 90C. I used the laptop with this CPU temp for more than 3 years; last year I changed to another laptop because the chasis of this HP DV6 is broken in too many places already; otherwise, it's still running fine. – imin Aug 15 at 16:54
  • @user41855 yes, this is above normal temperatures for idle. – Kalamalka Kid Aug 15 at 17:34
  • 2
    @user41855 if your cpu temp is that high at startup then there is definitely something wrong with your cooler. You need to fix it or your cpu won't last much longer, and it'll be extremely slow in the meantime. – Turksarama Aug 16 at 12:51
17

You're well into the range that can cause permanent damage to the CPU. Most consumer electronics (including CPUs) are not designed to operate above 85 °C for any extended period of time, and most will actually shut down when they get over about 100-105 °C. Provided you have a working (and properly sized) cooling system and are not somewhere with unusually high temperatures to begin with (40 °C or higher), you should not be seeing temperatures that high no matter how hard you push the CPU.

However, I'm inclined to believe something is wrong with your system due to that insane discrepancy between reported package and core temperatures. In particular, the possibilities that come to mind are:

  • The sensor isn't being read correctly and the core temperature is actually much lower. This is the best possible case, and it's easy to check (try a handful of other tools for reading these sensors. Everything else reporting similar temperatures does not rule this out though, because the drivers being used to make the reading may be bad (you can check that case by booting into a live Linux environment and seeing what it says the temperatures are. I know 100% for certain that Linux reports the temperature correctly on that model of CPU because the sensor interface the CPU provides has been around since the AMD K10 days and is very well supported by Linux).
  • The sensor isn't being read correctly, and the package temperature is actually much higher. This is extremely unlikely, because for it to be the case you have to have somehow managed to run a CPU with a 20 W TDP so hard that it got that hot. The only possibilities I can think of that would allow for that are running with no cooling system at all or running in an environment that was already unlivably hot for humans.
  • Something is wrong with one of the temperature sensors. It is not very likely, but it is still possible. If you've eliminated the two above possibilities, then this one can be checked by using a (good) infrared thermometer or (real) thermal camera to get an estimate of the temperature of the junction between the heat-sink and the IHS. If that reads back at close to 115 °C, then the package temperature sensor is bad (and something else is probably wrong with your CPU). If it reads back at close to the 66 °C being reported by the package temperature sensor, then either something is wrong with the core temperature sensor or the next (worst case) possibility is the case.
  • Something is physically wrong with the thermal junction inside the package between the IHS and the CPU die. This is the absolute worst case scenario, as it means your chip is essentially useless (because you quite simply cannot cool it well enough for it to be practically usable. This is also astronomically unlikely (actually, it's beyond astronomically unlikely, it's even beyond the unlikelihood of a SHA-256 hash collision with two randomly chosen files), but it's still technically possible. There's unfortunately no practical way to check this one if you've eliminated all the other possibilities, because delidding the CPU to manually check will make it irrelevant (and also require use of a completely different cooling system).
| improve this answer | |
  • At the startup the temperatures range between 90-95°C. So it is generally high!! – user41855 Aug 14 at 16:30
  • 2
    You're practically contradicting yourself when talking about the thermal junction. Failure of the thermal junction is either statistically possible or statistically impossible. If it's statistically possible, then your statement that "it's even beyond the unlikelihood of a SHA-256 hash collision with two randomly chosen files" is incorrect. If it's statistically impossible, then it's wrong to write of it as being a genuine possibility rather than a theoretical curiosity, and in particular, it's wrong to say that "There's unfortunately no practical way to check this one." – Tanner Swett Aug 15 at 15:36
  • @TannerSwett Just because something is statistically impossible does not mean it can't happen or doesn't happen. A human being struck by a meteorite is statistically impossible, but it has happened at least once. I'm erring on the side of completeness here, hence my inclusion in the list of possibilities. I also stand by my statement that there is no practical way to check it because you have to delid the CPU to do so, and doing that inherently destroys the potentially suspect thermal junction. – Austin Hemmelgarn Aug 15 at 21:35
  • 1
    Well, if you're erring in the side of completeness, how come you didn't mention the possibility that an evil wizard placed a curse on the CPU? If, as you say, a thermal junction failure is less likely than a SHA-256 collision in a single random trial, then the idea that the problem is caused by a thermal junction failure is certainly more ridiculous than the idea that the problem is caused by an evil wizard. – Tanner Swett Aug 16 at 4:08
  • 3
    Put another way, you're simultaneously claiming that (a) a thermal junction failure is so unlikely that, if it happened, then it's literally the most unusual thing to have ever occurred in the entire course of human history, and (b) a thermal junction failure is likely enough that the possibility ought to be considered. That's a contradiction. – Tanner Swett Aug 16 at 4:10
6

CPU core temperatures of more than 110 degrees is too high and make the processor stop working. If really temp is over 110 degrees system will likely crash and there could be nasty situation. HWMonitor is showing incorrect values.

Go to your BIOS and check temperatures there, and if BIOS also says nothing this could be indicative that your Motherboard dosen't have any temperature sensors, thus explaining why HWMonitor is showing strange values.

If the temperature is high, then you can check the airflow, clean the dust in casing and components and check if the fan is moving properly. You can use Core Temp as the other answer suggested.

| improve this answer | |
  • How would the motherboard be able to read core temperatures if that capability isn't already part of the CPU? – Mark Ransom Aug 14 at 16:49
  • How do you know it's showing incorrect values? – BlueRaja - Danny Pflughoeft Aug 14 at 18:33
  • "110 degrees is too high and make the processor stop working" -- aren't there options in the BIOS to disable this? Perhaps they're off? – Captain Man Aug 14 at 19:02
  • 1
    @CaptainMan 110 degrees is too high, as in the processor will gradually change its physical properties until it no longer performs calculations for you. – wizzwizz4 Aug 14 at 19:26
  • 1
    @CaptainMan BIOS can provide reading, but to decrease temp we need to make measures. Also from BIOS we can't get reading while gaming or doing other heavy things so we need a reliable software. – Wasif Hasan Aug 15 at 5:31
1

Your CPU temperature is reaching its very maximum. I think, any PC component having temperature over 100C is too much for extended period of time. It is recommended to keep it under 80C in the long run, 60-something is even better.

I always recommend to remove the original cooling fan of the CPU (which usually give you when you buy it) and buy a better cooler with bigger heatsink. I bought CoolerMaster HYPER TX3i, which is a cheap (under $30), but an effective one.

For my i3-9100F temperature with the original heatsink cooler: Idle - 65C, Load - 95 - 100C; after applying CoolerMaster: Idle - 30 - 42C, Load - 50 - 60C.

| improve this answer | |
  • Are these numbers correct? My i9 is ~35C idle and ~75C at full core load. At ~90C it throttles down and a while later (I don't have an exact number, but a little over 100) it shuts down. – Blindy Aug 17 at 15:58
  • i9 is more powerful, than i3, the temperature difference comes from this fact, I assume. – Ryksa Aug 18 at 9:17
1

When a CPU reaches 100+ degrees, it should be restarted. If it still doesn't' resolve the issue, you might want to reset your PC or delete some apps. Another reason might be the heatsink or the cooler's problem. If you don't want to get new hardware, I recommend you turn the CPU voltage a bit lower. If I am correct, your CPU is not unlocked, so you can't really adjust your clock speed, but beware when turning your voltage down, turn it down bit by bit. If you notice instabilities, revert the changes.

| improve this answer | |
1

The AMD A10-7300 has a maximum operating temperature of 102°C so your system should be shutting itself down to prevent damage.

The iGPU (Radeon R6) temp of 66°C appears to be much more believable. Since this matches the "Package Temp" then I would surmise that your actual temp is 66°C since the CPU cores and iGPU are in the same housing.

Given the fact that your cores are all running well above the stock 1900 MHz and one of them running near the max 3200 MHz turbo, it tells me that the temperatures are not affecting performance. If you're temperatures were too high then your CPU would throttle down to prevent damage.

You can download other temp monitoring software and see if you get similar results. Also, if your BIOS reports temps then you should check temps in your BIOS since that should be the most accurate reading.

| improve this answer | |
0

The temperature you told and I am seeing from the software is actually not that high! Because it can happen if you play high-end games or video editing software and benchmark software.

Although it's generally not an ideal temperature for a computer, but if the temperature stays while you do simple working and light or mid high works in your computer then it can be your cooling system. Which is failing!

You can use the software CPU-Z for testing your computer's temperature and then reask the question. Because sometimes software can be faulty, but CPU-Z is one of the best!

The ideal temperature for a normal working computer is between 62-66 degree Celsius. Anything above can be bad for your motherboard and its components if not for the processor itself.

You can download CPU-Z from here: CPU-Z

| improve this answer | |
  • 3
    CPUID HWMonitor (which the screenshot is taken from) is made by the same people as CPU-Z. – Lyndon White Aug 14 at 20:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.