Real Temp and Core Temp reporting different temperatures from HWMonitor??

[CAT-THE-FIFTH]

I noticed that HWMonitor is reporting 10C higher temperatures for my Xeon E3 than both Core Temp and Real Temp??

What is correct?? The motherboard is a Gigabyte GA-H67N-USB3-B3.

 

[Rooster212]

Either could be correct, or both could be wrong. Software temperature readings are never going to be completely accurate anyway. Maybe take the average of the two, I don't know what else to recommend, apart from maybe try some of Gigabyte's own programs, but I don't think that they have their own monitoring program.

 

[CAT-THE-FIFTH]

Thanks!

HWiNFO64 seems to show the same temperatures as Core Temp and Real Temp too.

Core Temp says the Tj. max for the Xeon E3 1220 is 94C and HW Monitor says 102C. What is correct??

 

[Bacon?]

They seem dangerously high!

 

[CAT-THE-FIFTH]

The Tj. max is not the reported temperature,but supposedly that is what is used to calculate the temperature seen. AFAIK,for a Core i5 2500K it is 98C,but I can't find much information for a Xeon E3.

 

[ubersonic]

Core/Real temp are correct, had this issue with my E3 too, as a general rule for temps always trust real/core temp over HWMonitor.

Which E3 is it? if its a V2 then it will have the same Tj Max as the equivalent IB CPU and if its not it will have the same Tj Max as the equivalent SB CPU.

 

[passey89]

Core/Real temp are correct, had this issue with my E3 too, as a general rule for temps always trust real/core temp over HWMonitor.

Which E3 is it? if its a V2 then it will have the same Tj Max as the equivalent IB CPU and if its not it will have the same Tj Max as the equivalent SB CPU.

yeh right core temp says at idle my temps at 15c.

 

[ubersonic]

yeh right core temp says at idle my temps at 15c.

I didn't say they were infallible, CT will misreport sometimes when it cannot identify the CPU and RT doesn't always pick the correct TJ Max, just that 9/10 when you have a clash between HWM and Coretemp+Realtemp its HWM at fault.

But as long as he is using up to date versions then wiht his CPU its HWM thats wrong.

 

[CAT-THE-FIFTH]

Thanks!

Just a bit of an information the CPU is an undervolted Xeon E3 1220,with Turbo switched off and max TDP set to 65W in the BIOS. I am using it in a SFF PC.

The thing is why I am fretting over this,if I use anything apart from HWMonitor I get acceptable temperatures when the CPU is under load,ie,around 65C when playing a game wich stresses all 4 cores. However,HWMonitor is showing nearer to 80C. The thing is the heatsink is warm to the touch but not scorchingly hot.

You can see HWMonitor is reporting 8C to 10C higher than the other packages.

Real Temp.

Core Temp

Tj. Max is set at 94C.

HWmonitor

Tj. Max is set at 102C.

HWiNFO64

 

[Mike_P]

Well HWMonitor reads 8°C higher because it has TJmax set 8°C higher. Distance to TJmax is what the sensor reads so you need to know what the TJmax is for your cpu.

I vaguely recall the guy (or it might of been someone on their forum) who did RealTemp spent time drilling into his cpus using other sensors and stuff to try and make RealTemp as accurate as possible and there was a big debate on what the TJmax for various cpus were. This was all awhile back perhaps the info is more readily available thesedays I've no idea.

 

[CAT-THE-FIFTH]

Thanks!

It seems even for the Xeon E3-1230 the Tj. Max is also 94C according to Real Temp:

http://www.*******.com/forums/index.php?topic=108402.0

For the SB Core i5 it is 98C from what I can find. However,for the Xeon E3 there is no information I can find anywhere and I even tried to have a look on the Intel website too.

The Xeon E3 CPUs tend to be rated for a slightly lower maximum temperature IIRC.

 

[Curious]

Some time back I think that there was a lengthy discussion about this, and I think either intel or amd (someone linked a mailresponse) came out saying that the lower temps seems to be more erronously reported, as in when the chip is pulling very little. But as you press the chip more the higher temperatures tend to be more correct.

So even if you get whacky lows (does it really matter much that it reports 15 degrees), the highs tend to be something you can use and take serious if it becomes too high.

 

[ubersonic]

The Xeon E3 CPUs tend to be rated for a slightly lower maximum temperature IIRC.

With the exception of the Xeon E3 CPUs whos model numbers end in 5, they have no GPU, hence why the TDP is lower than their i5/i7 counterparts, from what I have read the T-CASE temp is for the entire package and so shows lower for the chips without GPUs as their cores will reach the same temps as the ones in a CPU with a GPU but at higher package temp, if that makes sense.

For the SB Core i5 it is 98C from what I can find. However,for the Xeon E3 there is no information I can find anywhere and I even tried to have a look on the Intel website too.

If its the same as the IB E3's which I would expect it to be then the TJ.Max will be the same as the chips i5/i7 counterpart, its only the TDP that changes and thats only on Xeons with no GPU.


EDIT:

/Selffacepalm, I don't know how the hell I forgot to ask this earlier, have you always been using this CPU since before installing HWMonitor? or did you upgrade it afterwards? if you upgraded the CPU you will need to delete the hwmonitorw.ini file and let it generate a new one on next use as it will have the TJ.Max for your old CPU set in the config.

 

[CAT-THE-FIFTH]

Thanks again for the replies everyone!

EDIT:

/Selffacepalm, I don't know how the hell I forgot to ask this earlier, have you always been using this CPU since before installing HWMonitor? or did you upgrade it afterwards? if you upgraded the CPU you will need to delete the hwmonitorw.ini file and let it generate a new one on next use as it will have the TJ.Max for your old CPU set in the config.

I did that and now the temperatures agree!! Thank you very much!!

I upgraded the CPU from a Core i3 2100 but HWMonitor was installed before this.