Real Temp - New temp program for Intel Core processors
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Real Temp - New temp program for Intel Core processors
I'm going to try this new program for temps. The thread also has an (another!) explanation for why these Core2's temperature readings are a source of much debate, especially Wolfdale........
http://www.xtremesystems.org/forums/sho ... p?t=179044
http://www.xtremesystems.org/forums/sho ... p?t=179044
Well, this gay is guessing that TjMax for my processor (E4400) is 85C so RealTemp shows 27/27C at idle. CoreTemp and CPUID Hardware Monitor are both guessing 100C and accordingly shows 42/42C.
I suppose, you will never know the actual core temp for sure, but never mind, they all tell me the same, that I have 58C left to TjMax, and that's enough.
I like CPUID HWM because I get all temperatures (CPU,GPU,HDD) with one program, without extra functions that I never use. Besides, it also shows Max temperatures since start up.
I suppose, you will never know the actual core temp for sure, but never mind, they all tell me the same, that I have 58C left to TjMax, and that's enough.
I like CPUID HWM because I get all temperatures (CPU,GPU,HDD) with one program, without extra functions that I never use. Besides, it also shows Max temperatures since start up.
I also have a E4400 and the Tjmax is definately 85ºC. 34/35 at idle, open test bed using Intel box cooler.
The revision is L2.
M0 revision might be 100ºC
A simple test I do is resume from S3.
If temperatures are closer to ambient then Tjmax assumed is correct, if temperatures are 15ºC higher than ambient, then Tjmax is wrong.
The revision is L2.
M0 revision might be 100ºC
A simple test I do is resume from S3.
If temperatures are closer to ambient then Tjmax assumed is correct, if temperatures are 15ºC higher than ambient, then Tjmax is wrong.
It's not 85C, and it's not 100C either (at least not generally - different CPUs all have different values).
I have an E4300.
When I turn on the computer in the morning (ambient kept at 20-21C), the temperature is about 27-28C. It gradually increases until it hits 36-37C.
Both values are off. Subtracting 15 from 27-28 would make for impossible temperatures, and not subtracting anything from 36-37 would make for really high idle temps in a decently-ventilated case with a bloody Ninja on the CPU.
So, I'd say it's something inbetween - Tj would be about 92.5C. That would make startup temp equal to ambient, and later on, as everything inside the case heats-up, the temperature would increase to about 30C, which is much more believable.
Or it's not linear.
Edit: Silly me. It's all explained there, with pictures showing differences between E8xxx and E6xxx (and E4xxx should also be different).
I have to try this app.
I have an E4300.
When I turn on the computer in the morning (ambient kept at 20-21C), the temperature is about 27-28C. It gradually increases until it hits 36-37C.
Both values are off. Subtracting 15 from 27-28 would make for impossible temperatures, and not subtracting anything from 36-37 would make for really high idle temps in a decently-ventilated case with a bloody Ninja on the CPU.
So, I'd say it's something inbetween - Tj would be about 92.5C. That would make startup temp equal to ambient, and later on, as everything inside the case heats-up, the temperature would increase to about 30C, which is much more believable.
Or it's not linear.
Edit: Silly me. It's all explained there, with pictures showing differences between E8xxx and E6xxx (and E4xxx should also be different).
I have to try this app.
Tested.
RealTemp says my E4300 (2.4 GHz, 1.1V) has a Tj of 85C. CoreTemp puts it at 100C.
When I booted up into Windows, CoreTemp claimed 30C for the cores. RealTemp obviously said 15C, which is physically impossible. So, I clicked on the ++ radiobutton immediately and got 22C... Which is how it should be, I think - ambient is around 20-21C.
After some time idling, it's 37C in CoreTemp, 28C in RealTemp.
Ran a little bit of Orthos small FFT, just for a minute. CT=55C, RT=42C.
That's it! I'm 99.9% sure that RT is correct. My CPU is undervolted, there should be plenty of airflow, I have a rev.A Ninja with a 1000 rpm 100 mm Kaze-Jyu... The temps given by RT are far more believable than those given by CT.
Forget about CoreTemp, this is the real deal!
RealTemp says my E4300 (2.4 GHz, 1.1V) has a Tj of 85C. CoreTemp puts it at 100C.
When I booted up into Windows, CoreTemp claimed 30C for the cores. RealTemp obviously said 15C, which is physically impossible. So, I clicked on the ++ radiobutton immediately and got 22C... Which is how it should be, I think - ambient is around 20-21C.
After some time idling, it's 37C in CoreTemp, 28C in RealTemp.
Ran a little bit of Orthos small FFT, just for a minute. CT=55C, RT=42C.
That's it! I'm 99.9% sure that RT is correct. My CPU is undervolted, there should be plenty of airflow, I have a rev.A Ninja with a 1000 rpm 100 mm Kaze-Jyu... The temps given by RT are far more believable than those given by CT.
Forget about CoreTemp, this is the real deal!
Here is a "How to measure CPU Temps" Guide.
http://www.tomshardware.co.uk/forum/221 ... ture-guide
Unfortunately there is quite much to read and do.
But the new link (and "how to") provided in the first post seems to be better.
http://www.tomshardware.co.uk/forum/221 ... ture-guide
Unfortunately there is quite much to read and do.
But the new link (and "how to") provided in the first post seems to be better.
Is it really necessary to know the exact core temperatures?
I think the most important value is precisely that reported by the on core sensors, margin to TjMax (throttling point).
No guessing and no calculation required. Perhaps everybody should start using that value when comparing CPU temperatures?
I think the most important value is precisely that reported by the on core sensors, margin to TjMax (throttling point).
No guessing and no calculation required. Perhaps everybody should start using that value when comparing CPU temperatures?
not really.Matija wrote:Both values are off. Subtracting 15 from 27-28 would make for impossible temperatures, and not subtracting anything from 36-37 would make for really high idle temps in a decently-ventilated case with a bloody Ninja on the CPU.
i have a well ventilated case, with an e4400 with a ninja on top, idling at 43-44 degrees.
when i let it sit overnight, i woke up this morning to see the lowest temp i have ever seen on it: 39 degrees C.
you probably have a pretty low ambient temp, and the case itself is probably sitting in a fairly well ventilated room. i have a moderate ambient temp, and the case is sitting in a slightly enclosed bigger case. (with the other computer it shares the case with on and active, the idle temp goes to 46-48... with both of them under load, around mid 60's)