HWiNFO64 sensor alerts

Mobomonkey

Distinguished
Sep 19, 2013
116
0
18,680
I'm running an Intel core i3-4130 Haswell on an ASRock B85M Pro4 Mobo using G.Skill Ripjaws 8Gb DDR5 1600MHz RAM with a Sapphire R9 270X GPU. W7 OS with HDD's
I've just installed HWiNFO64 to monitor my GPU temps primarily, but I now see that according to it's sensor readings my CPU cores are running at as much as 83 deg C, tho my Mobo control panel sensor was cooling from 65 deg C when I accessed it. I figure that maybe the Mobo CP figure is whole chip case temp. I've got an Alpine 11 Pro Rev.2 cooler fitted which looks substantially better than the stock cooler that came with the CPU, and I was very careful to fit it well. I can't easily find out what CPU core temps I should worry about with this model. Also, the HWiNFO64 allows me to set min and max core temps to issue a warning sound. So, what max temp should I set as the warning trigger and should I set a 'times sampled' figure of more than 1? Thanks.
 
Solution
Yes, it's Tcase.

The Thermal Control Circuit (TCC) activation temperature is 100°C. When this temperature is exceeded this circuit kicks in to throttle the CPU. The TCC temperature can actually be lower or higher than 100°C and is calibrated at the fab on a part-by-part basis.

Each processor execution core has an on-die Digital Thermal Sensor (DTS) that detects the core's instantaneous temperature with a Taccuracy of ±5°C. HWiNFO64 shows the DTS temperature for each core.

Ideally you should try to keep the core temperature below 70°C to obtain maximum CPU life.


Intel specifies 72°C as the maximum temperature allowed at the processor Integrated Heat Spreader (IHS) for the Intel Core i3-4130 Processor.
 


Thanks ko888. Is that the Tcase figure you're using? I always thought that was the operating temp and not a max figure. Also, as I alluded to in my question, there's a difference between 'cores' temps and overall temp. Plus the 'times sampled' thing adds complexities too. I'm not sure it's as simple as setting my sensor alerts to the Tcase figure, besides, when I first did that, they were alerting me all the time. :)
 
Yes, it's Tcase.

The Thermal Control Circuit (TCC) activation temperature is 100°C. When this temperature is exceeded this circuit kicks in to throttle the CPU. The TCC temperature can actually be lower or higher than 100°C and is calibrated at the fab on a part-by-part basis.

Each processor execution core has an on-die Digital Thermal Sensor (DTS) that detects the core's instantaneous temperature with a Taccuracy of ±5°C. HWiNFO64 shows the DTS temperature for each core.

Ideally you should try to keep the core temperature below 70°C to obtain maximum CPU life.
 
Solution


Thanks ko888. It's obvious you know your onions. So, given that I've fitted a much bigger and better cooling fan than Intel provided, and that I always allow my games or raptr to set the specs, why is it running the cores so hot? I'm intending to upgrade to an i7 eventually, when I can afford it, but maybe I should get an i5 sooner if that would help. I guess I will have to manually adjust some key game specs down in the meantime. Given that the CPU cores are frequently running at 80-83 deg C at present, is it possible to guess at how quickly it will wear my CPU out if I left it as is?
 
You need to make sure that the voltage supplied to the CPU (i.e. Vcore) is not above Intel's spec when the CPU is running at full processing load. When the Vcore is too high your CPU temperature will also be excessive.

I think the normal Vcore for the Intel Core i3-4130 when at full load using stock 3.4 GHz clock speed is 1.106 Volts. You can check your CPU Vcore using the freeware utility CPU-Z.
 


Thanks ko888. I checked that, and it reads 1.07v max, so I'm guessing that's not the cause. As stated in my original post, I seem to have a different temp reading from my Mobo sensors. I checked with ASRock and they're adamant that their sensors are accurate and measure the Tcase temp. I'm gonna try contacting Intel to see what sensors HWiNFO64 might be accessing. I chose HWiNFO after some research, and it seems quite well recommended, but I'm having doubts as to the veracity of it's figures. Thanks again, I appreciate your advice. :)
 
HWiNFO64 reads from various sensors. On my Intel i7-4790K it even reads and displays the on-die Digital Thermal Sensor (DTS) temperature. The maximum temperature reported by the DTS for any core seems to match the Tcase temperature.
HWi_NFO64_Temperatures.jpg
 


Thanks ko888. If I assume that your figures are after a full load session, ie gaming, then I guess you're demonstrating an example of HWiNFO working correctly and therefore their figures are trustworthy. I note that your CPU has a Tcase of 74deg C and that your HWiNFO figures are only just within that max, so I guess there'd be little point in getting an i5 then! :) I still have some reservations because HWiNFO is only software and therefore as susceptible to the usual glitches and jinx's we're used to with all software. The reason why I mentioned the 'times sampled' in my question is that as comprehensive as HWiNFO's info is, I can't find just how long any of the cores were running at the max temps for. I can easily imagine a situation where my cores might well exceed the Tcase figure for micro seconds or even for 2 or 3 seconds during an intensive video rendition. Whether this constitutes excessive wear and tear I don't know and it's one of the reasons I've e-mailed Intel about this. The other reason is that my Mobo's CPU temp readings are significantly lower than HWiNFO's and may be using the same sensors. If so, then it's only software that can explain the difference. I'm looking forward to receiving Intel's response to my questions. Anyway, I've manually adjusted all my game specs downward and stopped playing Crysis 3 for now. :)
 

TRENDING THREADS