I know that in Bios it will say at what temp the CPU is running at…as well as another temp. But is there anyway to check these without going to Bios set up?
Try licking your cpu … if tongue burned … too hot … close down computer …
No, serious now … some brands of motherboards have a small software add-on that can display most critical info about your machine … and enables you to tweak your CPU and/or ram settings
I don’t know how to check your CPU temperature, but I do know a freeware program to check your hard drive temperature. It’s called “HDD Temperature”. You can read a description of it and download it from this website.
It’s an old, old thread, but I thought I’d bring it back to life rather than start a new one.
Our computer has crashed a few times in the past week. I opened it up, dusted it out–boy, did it ever need that!–and even cleaned off the fan to the power supply. I’m currently running it with the side panel off, which seems to solve the problem for now, but I’m worried about it getting even dirtier over time. I’d like to make it last as long as I can. There doesn’t seem to be a very strong airflow coming from the power supply, so I wonder if that may be the problem. I’ve never had reason to feel the airflow from the power supply before, so I’m not sure what normal or acceptable are.
Are there any new programs for monitoring computer heat? A few have been listed in this thread, but I’d like to know what other people are using or have read about.
The computer is an Acer M35 running Windows XP.
I use speedfan, it works, I guess.
For some weird reason my laptop can run just about anything, like 720p movies, but when I run a flash video for 5 mins it melts down.
Mine’s 39C - is that normal?
Temperatures in computers are really a very, very complicated topic. First of all, having any temperatures “measured” (read: “shown”) by any software, will inevitably lead to questions like:
Basically, not knowing what the 39°C are, let’s assume it’s CPU temperature. 39°C would be very, very far away from critical temperatures for any known CPU type. Most CPUs can stand something like ~80°C to 100+°C, depending on model.
But we have 2 problems here:
Measurement of temperatures is hard enough if you know what you are doing. Think of a professional system engineer at a big IT company, with access to good measurement equipment and the necessary specifications etc. For the CPU typically he will install a thermocouple in a groove he machined into the heatspreader of the CPU, and glue the thermocouple in there. All software that reads some registers in the CPU (or even worse, especially on older CPUs, some onboard chip that uses a sensor near the CPU or the thermal diode in the CPU) will have a very low accuracy, or often even assuming things about sensor readings that are not true (like for some older intel CPUs read the thermal sensor and show this as “°C”). Especially if neither you nor the software is exactly knowing what they are doing (like a generic tool eg. speedfan). All it can do is present some raw data from some sensor - but it doesn’t know how accurate the sensor is, how the ma inboard is designed, or what modifications need to be done (offset or whatever) to transform this temperature into anything useful… So basically: don’t trust any temperatures too much. They can easily be off by dozens of Kelvin…
Even once you have an accurate temperature (or just don’t believe me and assume what you have is accurate) it is most of the time very hard to know which temperature is OK. Said professional system engineer will look into data sheets / specifications for the components like the CPU, and find one of many different “Maximum temperature” specifications. That could be max. ambient temperature (together with a requirement like so and so many cfm airflow if you are lucky), or maximum case temperature (which is the only thing he can measure accurately) or junction temperature and thermal resistance (basically the silicon temperature an a way how to calculate it from the case temperature and the known or measured power dissipation).
So why did I write all that techno babble up there? It is to suggest that you guys (doesn’t matter if laymen or enhusiasts) just avoid such temperature guessing games.
First of all most people should be best off buying a PC where you know it has been engineered to have all components within their thermal spec if your ambient is within reasonable limits (most PCs etc. assume max. 35°C). Preferrably a professional PC - even if they seem a bit more expensive and much more “boring” than consumer PCs. If you worry about stability and reliability, go professional. That’s one of the reasons why professionals go for professional stuff and not consumer stuff
If you insist on building yourself (because you think you know how to do it, and it’s so easy) then you just gotta take some risk, or oversize you cooling to be safe. And then just ignore all temperature readings and simply be content if the PC runs OK
Now… if you have your PC crash on you and you think it might be heat related… there are some things you can do:
Check for excessive dust in the PC and maybe do a preventive cleaning
Look at all fans to see if they are turning and none are stopped (assuming none are supposed to be shut off, but with home built PCs that is seldom the case)
Like with every issue: You need a way to duplicate it reliably. Duplicate, try one solution, try to duplicate again. If it’s gone, you know the last thing you did fixed the issue. For such heat related issues (if it’s the CPU) one thing you can try at home is using Prime95’s torture test (google it). Let it run (with closed PC case) for some hours. It should not restart, hang up, or produce any errors. For the graphics card you can try stuff like furmark (google).
Once you can duplicate, try to verify it’s really heat related. Open the case and have a fan blow cool air inside the system. Preferably turn the AC low and let it blow in the PC direction as well. If that makes the error go away you got some heat problem for sure.
When running those stress tests on CPU or graphics card the respective cooler/heatsink should get noticeably warmer. If it doesn’t, there could be something wrong with the heat transfer between the chip and the heatsink. Think of too much (and maybe even additionally dried out) thermal paste, no thermal paste (or other thermal interface material), or maybe even still some protective foil in between the cooler and the CPU.
So if any of you guys believe their PC issues (which can be duplicated) are heat related, try some of the hints above. Feel free to post any additional questions that arise, some people here might be able to give more insight… just don’t give in to those temperature guessing and evaluating games, please
Hint: A quick sanity check for temperatures is as simple as using your hand. If you can touch some surface as long as you want, then you know it’s < ~50°C. Those temperatures are not critical for anything in a PC by the way, even if it “feels hot”. If you feel discomfort / pain touching the surface but can touch it a while then it’s most probably ~50-60°C. Also no issue for anything in a PC. If it’s too hot to touch (you have to retract your hand) then it’s for sure hotter than 50°C. Still no need to worry, though: all chips inside a PC can operate safely at temperatures that a human hand will not tolerate. These “touch checks” are just to give you an idea how hot some component really is. If your software shows “graphics card has 185°C” and you can touch the graphics card heatsink easily, then you know it is far, far off (or there is something seriously wrong between chip and heatsink)